Huge-scale video surveillance systems have become essential in crime prevention and situation recording. Traditional surveillance systems relied on human monitoring of video streams, which often led to errors and difficulties in understanding events. Furthermore, locating specific scenes within recorded videos required extensive human investigation. To overcome these challenges of inefficiency, inconvenience, and potential risks, we propose an intelligent analysis scheme that utilizes abnormal behavior recognition and metadata retrieval algorithms to replace human monitoring. The proposed method consists of three stages: i) metadata generation through object detection and tracking, ii) abnormal behavior recognition, and iii) SQL-based metadata retrieval. By incorporating specific information such as object color and aspect ratio, our technique enhances the usability of retrieval. Moreover, our abnormal behavior recognition module demonstrates robust classification capabilities for activities such as pushing, violence, falling, and crossing barriers. The proposed method can be seamlessly deployed on both edge cameras and analysis servers, making it adaptable to various surveillance setups. This approach revolutionizes the traditional surveillance paradigm, enabling more efficient, reliable, and secure video monitoring and analysis.

Here we present a blockchain-backed token recognition system to reward the contributions that academics make to the scientific ecosystem. Recognition is important in science but current methods, systems and incentives are limited. Specifically, the traditional focus on narrow publication metrics means diverse contributions are not captured, while bias toward senior, established scientists is common. To tackle this challenge, we explore the potential of harnessing blockchain’s collaborative, decentralised and trust-brokering properties to develop a token reward system for use by research funders. Academics would be awarded tokens for undertaking common but vital tasks such as peer review, sitting on funding committees and submitting reports. These tokens would not be tradable or specifically monetisable but would serve as a validated record of scientific contribution. They would have value in professional recruitment and job placement, support grant and award applications, and inform performance appraisals and file reviews. Coordination and cooperation across multiple funding agencies in developing the platform would provide an opportunity to aggregate and standardise recognition, given academics often work with several funders. This system’s goals are to expand recognition metrics, promote efficiencies, improve the robustness of professional assessments and enable cross-funder collaboration, thereby optimising research processes and practices in a decentralised and democratised manner.

The recent advances in Human-Computer Interaction and Artificial Intelligence have significantly increased the importance of identifying human emotions from different sensory cues. Hence, understanding the underlying relationships between emotions and sensory cues have become a subject of study in many fields including Acoustics, Psychology, Psychiatry, Neuroscience and Biochemistry. This work is a preliminary step towards investigating cues for human emotion on a fundamental level by aiming to establish relationships between tonal frequencies of sound and emotions. For that, an online perception test is conducted, in which participants are asked to rate the perceived emotions corresponding to each tone. The results show that a crossover point for four primary emotions lies in the frequency range of 417–440 Hz, thus consolidating the hypothesis that the frequency range of 432–440 Hz is neutral from human emotion perspective. It is also observed that the frequency dependant relationships between emotion pairs Happy—Sad, and Anger—Calm are approximately mirrored symmetric in nature.

Motivated by the way animals react in nature, we introduce a fully-connected neural-network model, which incorporates the external pattern presented to the system as a fundamental tool of the recognition process. In this new scenario, in the absence of this external field, memories are not attractors inside basins of attraction, as in usual attractor neural networks, although basins may be created according to the external pattern, thus allowing storing a much larger number of memories. The key point consists in calibrating the influence of the external pattern such as to cancel the noise generated by those memories not correlated with the external pattern.We illustrate how this proposal works by including this new contribution in the standard Hopfield model, showing a significant increase in its recognition capacity (typically by a factor $10^{2}$). As an additional feature of this model, we show its ability to react promptly to changes in the external environment, a crucial attribute of living beings. This procedure can be applied to a wide variety of neural-network models to increase considerably their recognition and reaction capabilities.

The traditional standards employed for pain assessment have many limitations. One such limitation is reliability because of inter-observer variability. Therefore, there have been many approaches to automate the task of pain recognition. Recently, deep-learning methods have appeared to solve many challenges, such as feature selection and cases with a small number of data sets. This study provides a systematic review of pain-recognition systems that are based on deep-learning models for the last two years only. Furthermore, it presents the major deep-learning methods that were used in review papers. Finally, it provides a discussion of the challenges and open issues.

Myoelectric control is the cornerstone of many assistive technologies used in clinical practice, such as prosthetics and orthoses, and human-computer interaction, such as virtual reality control. Although the performance of such devices exceeds 90\% in controlled environments, myoelectric devices still face challenges in robustness to variability of daily living conditions. Within this survey, the intrisic physiological mechanisms limiting practical implementations of myoelectric devices were explored: the limb position effect and the contraction intensity effect. The degradation of electromyography (EMG) pattern recognition in the presence of these factors was demonstrated on six datasets, where performance was 13% and 20% lower in realistic environments compared to controlled environments for the limb position and contraction intensity effect, respectively. The experimental designs of limb position and contraction intensity literature were surveyed. Current state-of-the-art training strategies and robust algorithms for both effects were compiled and presented. Recommendations for future limb position effect studies include: the collection protocol providing exemplars of 6 positions (four limb positions and three forearm orientations), three-dimensional space experimental designs, transfer learning approaches, and multi-modal sensor configurations. Recommendations for future contraction intensity effect studies include: the collection of dynamic contractions, nonlinear complexity features, and proportional control.

Multivariate Time Series Classification (MTSC) is one of most important tasks in time series analysis, aiding in activities such as human motion recognition and medical diagnostics. Existing methods for MTSC do not explicitly model temporal differences and generalize the idea of temporal difference into a efficient temporal module. Additionally, existing methods are not yet able to capture cross-variable relationships well during network training. As a result, they are unable to achieve convincing feature representation, leading to suboptimal classification accuracy. In this paper, we propose a novel MTSC model called Temporal Difference and Cross-variate Fusion Network (TDCFN), which integrates a two-stream differential LSTM network and a cross-variate feature extraction network to enhance feature representation. TDCFN achieves superior classification accuracy by capturing dynamic temporal evolution and inter-variable relationships. The experimental results show that TDCFN can achieve competitive performance with state-of-the-art multivariate time series classification approaches.

The voice contains elementary social communication cues, conveying speech, as well as paralinguistic information pertaining to the emotional state and the identity of the speaker. In contrast to vocal-speech and vocal-emotion processing, voice-identity processing has been less explored. This seems surprising, given the day-to-day significance of person recognition by voice. A valuable approach to unravel how voice-identity processing is accomplished is to investigate people who have a selective deficit in recognising voices. Such a deficit has been termed phonagnosia. In the present chapter, we provide a systematic overview of studies on phonagnosia and how they relate to current neurocognitive models of person recognition. We review studies that have characterised people who suffer from phonagnosia following brain damage (i.e. acquired phonagnosia) and also studies, which have examined phonagnosia cases without apparent brain lesion (i.e. developmental phonagnosia). Based on the reviewed literature, we emphasise the need for a careful behavioural characterisation of phonagnosia cases by taking into consideration the multistage nature of voice-identity processing and the resulting behavioural phonagnosia subtypes.

To improve the tangerine crop yield, the work of recognizing and then disposing of specific pests is becoming increasingly important. The task of recognition is based on the features extracted from the images that have been collected from websites and outdoors. Traditional recognition and deep learning methods, such as KNN (k-nearest neighbors) and AlexNet, are not preferred by knowledgeable researchers, who have proven them inaccurate. In this paper, we exploit four kinds of structures of advanced deep learning to classify 10 citrus pests. The experimental results show that Inception-ResNet-V3 obtains the minimum classification error.

Satellite scene classification is challenging because of the high variability inherent in satellite data. Although rapid progress in remote sensing techniques has been witnessed in recent years, the resolution of the available satellite images remains limited compared with the general images acquired using a common camera. On the other hand, a satellite image usually has a greater number of spectral bands than a general image, thereby permitting the multi-spectral analysis of different land materials and promoting low-resolution satellite scene recognition. This study advocates multi-spectral analysis and explores the middle-level statistics of spectral information for satellite scene representation instead of using spatial analysis. This approach is widely utilized in general image and natural scene classification and achieved promising recognition performance for different applications. The proposed multi-spectral analysis firstly learns the multi-spectral prototypes (codebook) for representing any pixel-wise spectral data, and then based on the learned codebook, a sparse coded spectral vector can be obtained with machine learning techniques. Furthermore, in order to combine the set of coded spectral vectors in a satellite scene image, we propose a hybrid aggregation (pooling) approach, instead of conventional averaging and max pooling, which includes the benefits of the two existing methods but avoids extremely noisy coded values. Experiments on three satellite datasets validated that the performance of our proposed approach is much more accurate than even the deep learning framework for spatial analysis.

Optical “fingerprints” are widely used in chemometrics-assited recognition of samples of different nature. An emerging trend in this area is the transition from obtaining "static" spectral data to reactions occurring over time. The indicator reactions are usually carried out in aqueous solutions; in this study we have developed the reactions that occur in an organic solvent, which makes it possible to recognize fat-soluble samples. In this capacity, we used 5W40, 10W40 and 5W30 motor oils of 4 manufacturers, totally 6 samples. The procedure involved mixing of the dye, sample, and reagents (HNO3, HCl, or t-butyl hydroperoxide) in ethanolic solution in a 96-well plate and measuring absorbance or near-IR fluorescence intensity every several minutes during 20–55 min. The obtained photographic images were processed by linear discriminant analysis (LDA) and k-nearest neighbors algorithm (kNN). The discrimination accuracy was evaluated by using the validation procedure. Reaction of oxidation of a dye with nitric acid allowed to recognize all 6 samples with 100% accuracy by LDA. Merging data of 4 reactions that did not provide complete discrimination ensured an accuracy of 93% by kNN technique. The developed indicator systems have good prospects for the discrimination of other fat-soluble samples. Overall, the results confirm the viability of the kinetic-based discrimination strategy.

Visual signals are the upmost important source for robots, vehicles or machines to achieve human-like intelligence. Human beings heavily depend on binocular vision to understand the dynamically changing world. Similarly, intelligent robots or machines must also have the innate capabilities of perceiving knowledge from visual signals. Until today, one of the biggest challenges faced by intelligent robots or machines is the matching in stereovision. In this paper, we present the details of a new principle toward achieving a robust matching solution which leverages on the use and integration of top-down image sampling strategy, hybrid feature extraction, and RCE neural network for incremental learning (i.e., cognition) as well as robust match-maker (i.e., recognition). A preliminary version of the proposed solution has been implemented and tested with data from Maritime RobotX Challenge (www.robotx.org). The contribution of this paper is to attract more research interest and effort toward this new direction which may eventually lead to the development of robust solutions expected by future stereovision systems in intelligent robots, vehicles and machines.

Emotion recognition and social inference impairments are well-documented post-traumatic brain injury (TBI) yet the mechanisms underpinning these are not fully understood. We examined dynamic emotion recognition, social inference abilities, and eye fixation patterns between adults with and without TBI. Eighteen individuals with TBI and 18 matched non-TBI participants were recruited and underwent all three components of The Assessment of Social Inference Test (TASIT). The TBI group were less accurate in identifying emotions compared to the non-TBI group. Individuals with TBI also scored lower when distinguishing sincere and sarcastic conversations but scored similarly to those without TBI during lie vignettes. Finally, those with TBI also had difficulty understanding the actor’s intentions, feelings, and beliefs compared to participants without TBI. No group differences were found for eye fixation patterns and there were no associations between fixations and behavioural accuracy scores. This conflicts with previous studies and might be related to an important distinction between static and dynamic stimuli. Visual strategies appeared goal- and stimulus-driven, with attention being distributed to the most diagnostic area of the face for each emotion. These findings suggest that low-level visual deficits may not be modulating emotion recognition and social inference disturbances post-TBI.

Facial Emotion Recognition (FER) is crucial across psychology, neuroscience, computer vision, and machine learning due to the diversified and subjective nature of emotions, varying considerably across individuals, cultures, and contexts. This paper investigates the impact of histogram equalization, data augmentation and model optimization strategies on three well-known FER datasets using pre-trained VGG models. Additionally, this paper showcases the effectiveness of different regularization techniques, callbacks, and learning schedulers in enhancing model performance by conducting extensive experiments. The model evaluation is discussed in terms of the following metrics: Accuracy, Area Under the Receiver Operating Characteristic Curve (AUC-ROC), Area Under the Precision-Recall Curve (AUC-PRC), and Weighted F1 score. Notably, the fine-tuned VGG models exhibit state-of-the-art performance compared to more complex conventional transfer learning models resulting in accuracies of 100%, 95.92%, and 69.65% on the CK+, KDEF, and FER2013 datasets, respectively.

Integrating voice assistant technology can potentially enhance accessibility and user interaction in web-based platforms for geospatial data visualisation and processing. This paper presents an open-source development incorporating a voice virtual assistant into the BStreams web application platform, designed specifically for visualising geospatial data. The study aimed to mitigate the observed limitations of virtual assistants in understanding the semantics and context of user commands in the geospatial domain. A discourse framework specialised for geospatial data was developed using a structured prototype methodology. In addition, a survey assisted in identifying commonly used English terminologies for geospatial data interaction. The questionnaire results were subsequently assessed using the ChatGPT natural language processing model. The application used open-source frameworks, including Web Speech API, Leaflet, Mapbox geocoding, and JavaScript. This investigation indicates that voice virtual assistants can potentially enhance human-computer interaction for geospatial data visualisation and processing. However, challenges related to natural language understanding, accents, and domain knowledge were raised. The paper advocates for further research to refine the integration of voice technology in this domain.

Collision avoidance is an important feature in advanced driver-assistance systems, aiming at providing correct, timely and reliable warnings before an imminent collision (objects, vehicles, pedestrians, etc.). A co-simulation framework is proposed in this paper to address the design and evaluation of collision avoidances in a cyber-physical system. The co-simulation framework is supported on the interaction between SCANeR and Matlab/Simulink. From the best of authors’ knowledge, two main contributions are reported in this paper. Firstly, the modelling and simulation of virtual on-chip LIDAR sensors in a cyber-physical system (CPS) considering traffic scenarios is presented. The CPS is designed and implemented in SCANeR. Secondly, an obstacle recognition library with three specific Artificial Intelligence-based methods is also designed based on sensory information database provided by SCANeR. Three methods for collision avoidance detection are considered, i.e.; a multi-layer perceptron neural network, a self-organization map and a support vector machine. Finally, a comparison among these methods for detecting obstacles before different weather conditions is done with very promising results in terms of accuracy. The best results are achieved using the multi-layer perceptron in sunny and fog conditions, the support vector machine in rainy and self-organized map in snowy conditions.

Caring for dolphins is a delicate process that requires experienced caretakers to pay close attention to their behavioral characteristics. However, caretakers may sometimes lack experience or not be able to give their full attention, which can lead to misjudgment or oversight. To address these issues, a dolphin behavior analysis system has been designed to assist caretakers in making accurate assessments. The study utilized image preprocessing techniques to reduce sunlight reflection in the pool and enhance the outline of dolphins, making it easier to analyze their movements. The dolphins were divided into 11 keypoints using an open-source tool called DeepLabCut, which helped accurately mark various body parts for skeletal detection. The AquaAI Dolphin Decoder (ADD) was then used to analyze six dolphin behaviors. To improve behavior recognition accuracy, the Long Short-Term Memory (LSTM) neural network was introduced. The ADD and LSTM models were integrated to form the ADD-LSTM system. Several classification models, including unidirectional and bidirectional LSTM, GRU, and SVM, were compared. The results showed that the ADD module combined with a double-layer bidirectional LSTM method achieved high accuracy in dolphin behavior analysis. The accuracy rates for each behavior were over 90%.

Due to the recent trend of unmanned economy, retail stores have gradually reduced their service and cashier manpower. The retail product recognition becomes one of key issues for unmanned shopping. Although the success of deep neural networks has made object recognition feasible in a variety of applications, it still struggles to perform well on a large number of classes of retail products. In this paper, we propose an improved ensemble learning-based approach for retail product recognition. In the proposed approach, the object classification network is first improved through feature extraction and block attention. An ensemble model is then built by integrating multiple network models and using loss selection as model weights. In the experiments, the feasibility of our ensemble learning-based approach has been evaluated through many production items. The results demonstrate the effectiveness of the proposed approach compared with previous retail product recognition methods.

Issues of fairness and consistency in Taekwondo poomsae evaluation have emerged owing to the lack of an objective evaluation method. This study proposes a three-dimensional (3D) convolutional neural network (CNN)-based action recognition model for the objective evaluation of Taekwondo poomsae. The model exhibits robust recognition performance regardless of variation in perspective by reducing the discrepancies between training and test images. The model uses 3D skeletons of the poomsae unit action collected using a full-body motion-capture suit to generate synthesized two-dimensional (2D) skeletons from the desired perspective. This approach aids in obtaining 2D skeletons from diverse perspectives as part of the training dataset and ensures consistent recognition performance regardless of the viewpoint. The model was trained using 2D skeletons projected from diverse viewpoints, and its performance was evaluated using various test datasets, including projected 2D skeletons and RGB images captured from various viewpoints. Comparison of the performance of the proposed model with that of previously reported action recognition models demonstrated the superiority of the model, underscoring its effectiveness in recognizing and classifying Taekwondo poomsae actions.

Identification of plants is a challenging task which aims to identify the family, genus, and species level according to morphological features. Automated deep learning-based computer vision algorithms are widely used for identifying plants and can help users to narrow down the possibilities. However, numerous morphological similarities between and within species make the classification difficult. In this paper, we tested a custom convolution neural network (CNN) and vision transformer (ViT) based models using the PyTorch framework to classify plants. We used a large dataset of 88K and 16K images for classifying plants at genus and species levels respectively. Our results show that for classifying plants at the genus level, ViT models perform better compared to CNN-based models ResNet50 and ResNet-RS-420, and other state-of-the-art CNN-based models suggested in previous studies on a similar dataset. The ViT model achieved top accuracy of 83.3% for classifying plants at the genus level. ViT models also perform better for classifying plants at the species level compared to CNN-based models ResNet50 and ResNet-RS-420, with a top accuracy of 92.5%. We show that the correct set of augmentation techniques plays an important role in classification success.

To achieve an acceptable level of security on the web, the Completely Automatic Public Turing test to tell Computer and Human Apart (CAPTCHA) was introduced as a tool to prevent bots from doing destructive actions such as downloading or signing up. Mobile devices have small screens, and therefore, using the common CAPTCHA methods (e.g. text CAPTCHAs) in these devices raises usability issues. To introduce a reliable, secure, and usable CAPTCHA that is suitable for mobile devices, this paper introduces a hand gesture recognition CAPTCHA based on applying Genetic Algorithm (GA) principles on Multi-Layer Perceptron (MLP). The proposed method improves the performance of MLP-based hand gesture recognition. It has been trained and evaluated on 2201 videos of the IPN Hand dataset, and MSE and RMSE benchmarks report the index values of 0.0018 and 0.0424, respectively. Comparison with the related works shows a minimum of 1.79% fewer errors, and experiments produced a sensitivity of 93.42% and accuracy of 92.27% – 10.25% and 6.65% improvement compared to the MLP implementation.

In offshore drilling, accidents such as gas invasion, overflow, and kicks are inevitable, and they can escalate into blowouts and other catastrophic events that result in casualties and substantial economic losses. Therefore, maintaining drilling safety requires accurate monitoring of gas invasion and overflow. The majority of overflow monitoring methods currently utilized at drilling sites are threshold-based. However, monitoring parameters acquired during actual drilling operations frequently contain noise signals, making it difficult for threshold-based methods to strike a balance between improving accuracy and minimizing false positives. In this paper, Pattern Recognition-based Kick Detection (PRKD) is proposed as a novel method for diagnosing overflow in offshore drilling. This method utilizes the overflow evolution process by integrating multi-phase flow calculations, data filtering theory, pattern recognition theory, the Bayesian framework, and other theoretical models. The PRKD effectively detects and monitors gas intrusion and overflow based on single parameters by analyzing the shape and wave characteristics of the curves. The case analysis demonstrates that the proposed method for monitoring drilling overflow achieves high precision while maintaining a low false positive rate. By combining advanced computational techniques with pattern recognition algorithms, the PRKD improves the accuracy and reliability of kick detection, enabling proactive responses to potential risks, protecting the environment and human lives, and optimizing drilling operations.

In recent years, with the popularity of smart mobile devices, the interaction between devices and users, especially in the form of voice interaction, has become increasingly important. If smart devices can understand more users' emotional states through voice data, more customized services can be provided for users. This paper proposes a novel machine learning model for speech emotion recognition, which combines convolutional neural networks (CNN), long short-term memory neural networks (LSTM), and deep neural networks (DNN), called CLDNN. To make the designed system can recognize the audio signal closer to the human auditory system does, this article uses the Mel frequency cepstral coefficients (MFCCs) of audio data as the input of the machine learning model. First, the MFCCs of the voice signal is extracted as the input of the model, and the feature values of the data are calculated using several local feature learning blocks (LFLB) composed of one-dimensional CNN. Because the audio signals are time-series data, the feature values obtained from LFLBs then input into LSTM layer to enhance the learning on time-series level. Finally, fully connected layers are used for classification and prediction. Three databases RAVDESS, EMO-DB and IEMOCAP are used for the experiments in this paper. The experimental results show that the proposed method can improve the accuracy compared to other related researches in speech emotion recognition.

In this paper we use quantum machine learning to detect and classify ocular diseases across age related macular degradation, cataract, diabetic, glaucoma, hypertension, and patological myopia categories versus a control group. We analyze fundus imagery from 1000 patients. Early findings indicate there may be benefit in terms of accuracy and loss function minimization of 2.07% and 1.979x respectively compared to a similar method implemented using traditional computers.

The growth in online child exploitation material is a significant challenge for European Law Enforcement Agencies (LEAs). One of the most important sources of such online information corresponds to audio material that needs to be analyzed to find evidence in a timely and practical manner. That is why LEAs require a next-generation AI-powered platform to process audio data from online sources. We propose the use of speech recognition and keyword spotting to transcribe audiovisual data and to detect the presence of keywords related to child abuse. The considered models are based on two of the most accurate neural-based architectures to date: Wav2vec2.0 and Whisper. The systems are tested under an extensive set of scenarios in different languages. Additionally, keeping in mind that obtaining data from LEAs is very sensitive, we explore the use of federated learning to have more robust systems for the addressed application, while maintaining the privacy of the data to LEAs. The considered models achieved a word error rate between 11% and 25%, depending on the language. In addition, the systems are able to recognize a set of spotted words with true positives rates between 82% and 98%, depending on the language. Finally, federated learning strategies show that they can maintain and even improve the performance of the systems when compared to centralized trained models. The proposed systems sit the basis for an AI-powered platform for automatic analysis of audio in the context of forensic applications within child abuse. The use of federated learning is also promising for the addressed scenario, where data privacy is an important issue to be managed.

This paper describes the successful application of the Transformer model used in the natural language processing and vision tasks as a means of processing the time series of signals from gyroscope and accelerometer sensors for the classification of human activities. The Transformer model is based on deep neural networks with many layers which can generalize well on signals. All measured signals come from a smartphone placed in a waist bag. Activity prediction is sequence-to-sequence, each time step of the signal is assigned a designation of the performed activity. Emphasis is placed on attention mechanisms, which express individual dependencies between signal values within a time series. In comparison with another recent result, the recognition precision was improved from 89.67 percent to 99.2 percent. The transformer model should in the future be included among the top options in machine learning methods for human activity recognition.

The multi-modal data fusion method based on IA-net and CHMM technical proposed is designed to solve the problem that the incompleteness of target behavior information in complex family environment leads to the low accuracy of human behavior recognition.The two improved neural networks(STA-ResNet50、STA-GoogleNet)are combined with LSTM to form two IA-Nets respectively to extract RGB and skeleton modal behavior features in video. The two modal feature sequences are input CHMM to construct the probability fusion model of multi-modal behavior recognition.The experimental results show that the human behavior recognition model proposed in this paper has higher accuracy than the previous fusion methods on HMDB51 and UCF101 datasets. New contributions: attention mechanism is introduced to improve the efficiency of video target feature extraction and utilization. A skeleton based feature extraction framework is proposed, which can be used for human behavior recognition in complex environment. In the field of human behavior recognition, probability theory and neural network are cleverly combined and applied, which provides a new method for multi-modal information fusion.

Retinopathy is a human eye disease that causes changes in retinal blood vessels that leads to bleed, leak fluid and vision impairment. Symptoms of retinopathy are blurred vision, changes in color perception, red spots, and eye pain. In this paper, a new methodology based on Convolutional Neural Networks (CNN) is developed and proposed to diagnose and give a decision about the presence of retinopathy. The CNN model is trained by different images of eyes that have retinopathy and those which do not have retinopathy. The performance of the proposed model is compared with the related methods of DREAM, KNN, GD-CNN and SVM. Experimental results show that the proposed CNN performs better.

Indoor scene recognition has great significance for intelligent applications such as mobile robots, location-based services (LBS) and so on. Wherever we are or whatever we do, we are under a specific scene. The human brain can easily discern a scene with a quick glance. However, for a machine to achieve this purpose, on one hand, it often requires plenty of well-annotated data which is time-consuming and labor-intensive. On the other hand, it is hard to learn effective visual representations due to large intra-category variation and inter-categories similarity of indoor scenes. To solve these problems, in this paper, we adopted an unsupervised visual representation learning method which can learn from unlabeled data with a Siamese Convolutional Neural Network (Siamese ConvNet) and graph-based constraints. Specifically, we first mined relationships between unlabeled samples with a graph structure. And then, these relationships can be used as supervision for representation learning with a Siamese network. In this method, firstly, a k-NN graph would be constructed by taking each image as a node in the graph and its k nearest neighbors are linked to form the edges. Then, with this graph, cycle consistency and geodesic distance would be considered as criteria for positive and negative pairs mining respectively. In other words, by detecting cycles in the graph, images with large differences but in the same cycle can be considered as same category (positive pairs). By computing geodesic distance instead of Euclidean distance from one node to another, two nodes with large geodesic distance can be regarded as in different categories (negative pairs). After that, visual representations of indoor scenes can be learned by a Siamese network in an unsupervised manner with the mined pairs as inputs. In order to evaluate the proposed method, we tested it on two scene-centric datasets, MIT67 and Places365. Experiments with different number of categories have been conducted to excavate the potential of proposed method. The results demonstrated that semantic visual representations for indoor scenes can be learned in this unsupervised manner. In addition, with the learned visual representations, indoor scene recognition models trained with the learned representations and a few of labeled samples can achieve competitive performance compared to the state-of-the-art approaches.

Gender recognition is trivial for physiotherapist, but it is considered a challenge for computers. The electromyography (EMG) and heart rate variability (HRV) were utilized in this work for gender recognition during the stepping exercise using a stepper. The relevant features were extracted and selected. The selected features were then fused to automatically predict gender recognition. However, the feature selection for gender classification became a challenge to ensure better accuracy. Thus, in this paper, a feature selection approach based on both the performance and the diversity between the two features from the rank-score characteristic (RSC) function in a combinatorial fusion approach (CFA) was employed. Then, the features from the selected feature sets were fused using a CFA. The results were then compared with other fusion techniques such as naive bayes (NB), decision tree (J48), k-nearest neighbor (KNN) and support vector machine (SMO). Besides, the results were also compared with previous researches in gender recognition. The experimental results showed that the CFA was efficient for feature selection. The fusion method was also able to improve the accuracy of the gender recognition rate. The CFA provides much better gender classification results which is 94.51% compared to Nazarloo's work (90.34%) and other classifiers.

The utilization of Pig Emotion Recognition (PER) driven by Artificial Intelligence (AI) promises to mitigate labor costs and alleviate stress among domestic pigs, thereby minimizing the need for consistent human intervention. Nevertheless, this research acknowledges the inherent limitations within the raw PER datasets, which often include irrelevant porcine features, hence impeding genuine progress in real-world evaluations. A significant proportion of PER datasets derive from sequential pig imagery obtained from video recordings, and a common pitfall in these studies is the unregulated shuffling of data. This lack of control can result in the overlap of data samples between training and testing groups, thereby yielding skewed experimental evaluations. To address these challenges, this paper introduces a novel solution in the form of the Semi-Shuffle-Pig Detector (SSPD) for PER datasets, with the intent to facilitate a less biased experimental output. By applying the SSPD, we can ensure that all testing data samples remain distinct from the training datasets, and any superfluous information from raw images is systematically discarded. This optimized method enhances the true performance of classification, providing unbiased experimental evaluations. Notably, our approach has led to a remarkable improvement in the Isolation After Feeding (IAF) metric by 20.2\% and achieved higher accuracy in segregating IAF and Paired After Feeding (PAF) classifications exceeding 92\%. This methodology, thereby, ensures the preservation of pertinent data within the PER system and eliminates potential biases in experimental evaluations. Consequently, it elevates the accuracy and reliability of real-world PER applications, resulting in a tangible positive impact on both pig welfare management and food safety standards.

This study leverages transformer-based models and focuses on data representation strategies in the named entity recognition task, including "single" (one sentence per vector), "merged" (multiple sentences per vector), and "context" (sentences joined with attention to context). Performance analysis reveals that models trained with a single strategy may not perform well on different data representations. A combined training procedure is proposed to address this limitation, using all three strategies to enhance the stability and adaptability of the model. The results of this approach are presented and discussed for various datasets for four languages (English, Polish, Czech, and German), demonstrating the effectiveness of the combined strategy.

This study aimed to investigate the possibility of using one-shot hyperspectral airborne images to recognize crops for an area with many small plots. The results showed that unsupervised clustering methods could classify crops with an accuracy of 80%, which improved to 90% when restricted to only grain crops, using a single airborne hyperspectral recording. However, additional layers such as NDVI, DTM, slope, and aspect did not improve classification accuracy. For comparison, the accuracy of clustering time series Sentinel-2 images with NDVI layers and DTM-derived data yielded an accuracy of: 74% ,Sentinel-2 time series 68% and single one registration before harvest - 39%. The results of the random forest classification were slightly less accurate due to a lack of sufficient reference data. However, it is challenging to verify the reported accuracy of crop recognition in the literature above 90% due to differences in analysis methodologies, reference data selection, pixel/object approaches, metric choice, and calculation formulas used.

Currently, the most important challenge in any assessment of state-of-the-art sensor technology and its reliability is to achieve road traffic safety targets. The research reported in this paper is focused on the design of a procedure for evaluating the reliability of Internet-of-Things (IoT) sensors and the use of a Cyber-Physical System (CPS) for the implementation of that evaluation procedure to gauge reliability. An important requirement for the generation of real critical situations under safety conditions is the capability of managing a co-simulation environment, in which both real and virtual data sensory information can be processed. An IoT case study that consists of a LiDAR-based collaborative map is then proposed, in which both real and virtual computing nodes with their corresponding sensors exchange information. Specifically, the sensor chosen for this study is a Ibeo Lux 4-layer LiDAR sensor with IoT added capabilities. Implementation is through an artificial-intelligence-based modeling library for sensor data-prediction error, at a local level, and a self-learning-based decision-making model supported on a Q-learning method, at a global level. Its aim is to determine the best model behavior and to trigger the updating procedure, if required. Finally, an experimental evaluation of this framework is also performed using simulated and real data

The content of this paper is a very significant contribution to the heritage regional reconstruction and conservation interventions. The results about the iconographic analysis of the Early-Renaissance frescoes of St. Leocadia Church, in Chaves, in the north-east of Portugal, were presented for the first time. This paper aims to study facial similarity analysis applied to identification of the personages represented in the 1511-1513 mural paintings of the apse of St. Leocadia Church, in Chaves Municipality (North of Portugal) and those ones in the oil paintings by the Proto-Renaissance Portuguese painter Nuno Gonçalves. In this work, it was explored the feasibility of face recognition technologies in answering many ambiguities concerning Manueline stylistic identity and iconography. A proposal was made to identify the characters, fundamental to the meaning of the frescoes. It has been experimentally proved on 7 characters of St. Leocadia Church’s paintings, in comparison with the referred old master portraits. It was used an in-house trained deep face recognition model to extract facial features from images of 15th and 16th century portraits to compare their degree of similarity. It was undertaken facial similarity analysis of the faces in the formidable portrait gallery of the 15th century Portuguese society represented in the Panels of St. Vicent, which has been subject of national and international research for 130 years. On the other, in the oil painting of St. Peter and St. Paul and Infanta St. Joana, of the same old master. The iconographic interpretation of the mural paintings of St. Leocadia, based in the deep facial features of the images, is plausible: a catechetical purpose; the ritual practices of royal ancestor worship in royal portrait apses of the churches; the Portuguese maritime expansion and the macro-imperial ideology of D. Manuel I.

The development of a novel and simple inhibition biosensor array for detection of water pollutants based on immobilized bacteria is the main goal of this work. A series of electrochemical measurements (i.e. cyclic voltammograms) were carried out on screen-printed gold electrodes with three types of bacteria, namely Escherichia coli, Shewanella oneidensis, and Methylococcus capsulatus, immobilized via poly L-lysine. For comparison purposes, similar measurements were carried out on bacteria samples in solutions,; also optical measurements (fluorescence microscopy, optical density, and flow cytometry) were performed on the same bacteria in both liquid and immobilized forms. The study of the effect of heavy metal ions (lead), pesticides (atrazine) and petrochemicals (hexane) on DC electrochemical characteristics of immobilized bacteria revealed a possibility of pattern recognition of the above inhibition agents in aquatic environment.

This paper presents preliminary research that investigates the relationship between the flow of a group of jazz musicians, quantified through multi-person pose synchronization, and their collective emotions. Building upon previous studies that measured the physical synchronicity of team members by tracking their body movements and measuring the difference in arm, leg, and head movements, we introduce a novel metric termed "team entanglement". We employ facial expression recognition to evaluate the musicians’ collective emotions. Through correlation and regression analysis, we establish that higher levels of synchronized body and head movements correspond to lower levels of disgust, anger, sadness and higher levels of joy among the musicians. Furthermore, we utilize a Convolutional Neural Network (CNN) based deep learning model to predict the collective emotions of the musicians. This model leverages 17 body synchrony keypoint vectors as features, resulting in a training accuracy of 61.47% and a test accuracy of 66.17%.

In this paper, an intelligent blind guide system based on 2D LiDAR and RGB-D camera sensing is proposed, and the system is mounted on the smart cane. The intelligent guide system relies on 2D LiDAR, RGB-D camera, IMU, GPS, Jetson nano B01, STM32 and other hardware. The main advantage of the intelligent guide system proposed by us is that the distance between the smart cane and obstacles can be measured by 2D LiDAR based on Cartographer algorithm, thus achieving Simultaneous localization and mapping (SLAM). At the same time, through the improved yolov5 algorithm, pedestrians, vehicles, pedestrian crosswalks, traffic lights, warning posts, stone piers, tactile paving and other objects in front of the visually impaired can be quickly and effectively identified. Laser SLAM and improved yolov5 obstacle identification tests were carried out inside a teaching building on the campus of Hainan Normal University and on a pedestrian crossing on Longkun South Road in Haikou City, Hainan Province. The results show that the intelligent guide system developed by us can drive the wheels and omnidirectional wheels at the bottom of the smart cane, and give the smart cane a self-leading blind guide function like a "guide dog", which can effectively guide the visually impaired to avoid obstacles and reach the predetermined destination, and can quickly and effectively identify the obstacles on the way out. The laser SLAM speed of this system is 25~31FPS, which can realize the short-distance obstacle avoidance and navigation function both in indoor and outdoor envionments. The improved yolov5 helps to identify 86 types of objects, the recognition rate for pedestrian crosswalks and for vehicles are 84.6% and 71.8%, respectively; the overall recognition rate for 86 types of objects is 61.2%, and the obstacle recognition rate of the intelligent guide system is 25-26FPS.

Industrial radiography is a pivotal non-destructive testing (NDT) method that ensures quality and safety in a wide range of industrial sectors. Conventional human-based approaches, however, are prone to challenges in defect detection accuracy and efficiency, primarily due to the high inspection demand from manufacturing industries with high production throughput. To solve this challenge, numerous computer-based alternatives have been developed, including Auto-mated Defect Recognition (ADR) using deep learning algorithms. At the core of training, these algorithms demand large volumes of data that should be representative of real-world cases. However, the availability of digital X-ray radiography data for open research is limited by non-disclosure contractual terms in the industry. This study presents a pipeline that is capable of modeling synthetic images based on statistical information acquired from X-ray intensity distribution from real digital X-ray radiography images. Through meticulous analysis of the intensity distribution in digital X-ray images, the unique statistical patterns associated with the exposure conditions used during image acquisition, type of component, thickness variations, beam divergence, anode heel effect, etc., are extracted. The realized synthetic images were utilized to train deep learning models, yielding an impressive model performance with mean intersection over union (IoU) of 0.93, and mean dice coefficient of 0.96, on real unseen digital X-ray radiography images. This methodology is scalable and adaptable, making it suitable for diverse industrial applications.

Against the backdrop of the current carbon peaking and carbon neutrality policies, higher requirements have been put forward for the upgrading and construction of smart grids. Non-intrusive load monitoring (NILM) technology is a key technology for advanced measurement systems at the end of the power grid. It obtains detailed power information of the load, without the need for traditional hardware deployment. The key step to solve this problem is load decomposition and identification. This paper first utilized Long Short Term Memory-Denoising Autoencoder(LSTM-DAE) to decompose the mixed current signal on the household busbar and obtain the current signals of multiple independent loads that constituted the mixed current. Then, the obtained independent current signals were combined with voltage signals to generate multi-cycle colored Voltage-Current(VI) trajectories, which were color-coded according to the background. These colored VI trajectories with background colors formed a feature library. When the CNN network was used for load recognition, considering the influence of hyperparameters on recognition results, the BOA algorithm was used for optimization, and the optimized CNN network was employed for VI trajectory recognition. Finally, the proposed method was validated using the PLAID dataset. Experimental results showed that the proposed method exhibited better performance in load decomposition and identification.

Recently, deep learning models have been widely applied to modulation recognition, which have become a hot topic due to their excellent end-to-end learning capabilities. However, current methods are mostly based on uni-modal inputs, which suffer from incomplete information and local optimization. To complement the advantages of different modalities, we focus on the multi-modal fusion method. Therefore, we introduce an iterative dual-scale attentional fusion (iDAF) method to integrate multimodal data. Firstly, two feature maps with different receptive field sizes are constructed using local and global embedding layers. Secondly, the feature inputs are iterated into the Iterative Dual Scale Attention Module (iDCAM), where the two branches capture the details of high-level features and the global weights of each modal channel, respectively. The iDAF not only extracts the recognition characteristics of each specific domains, but also complements the strengths of different modalities to obtain a fruitful view. Our iDAF achieves a recognition accuracy of 93.5\% at 10dB and 0.6232 at full SNR. The comparative experiments and ablation studies effectively demonstrate the effectiveness and superiority of the iDAF.

Cyber-physical security is vital for protecting key computing infrastructure against cyber attacks. Individuals, corporations, and society can all suffer considerable digital asset losses due to cyber attacks, including Data loss, theft, financial loss, reputation harm, company interruption, infrastructure damage, ransomware attacks, and espionage. A cyber-physical attack harms both digital and physical assets. Cyber-physical system security is more challenging than software-level cyber security because it requires physical inspection and monitoring. This paper proposes an innovative and effective algorithm to strengthen Cyber-Physical Security (CPS) with minimal human intervention. It is a Human Activity Recognition (HAR)-based approach where a GoogleNet-BiLSTM network hybridization has been used to recognize suspicious activities in cyber-physical infrastructure perimeter. The proposed HAR-CPS algorithm classifies suspicious activities from real-time video surveillance with an average accuracy of 73.15%. It incorporates Machine Vision at the IoT Edge (Mez) technology to make the system latency tolerant. Dual-layer security has been ensured by operating the proposed algorithm and GoogleNet-BiLSTM hybrid network from a cloud server, which ensures the security of the proposed security system. The innovative optimization scheme makes it possible to strengthen cyber-physical security with $4.29 per month only.

Smart devices such as smartphones, smartwatches, etc. are promising platforms that are being used for automatic recognition of human activities. However, it is difficult to accurately monitor complex human activities due to inter-class pattern similarity, which occurs when different human activities exhibit similar signal patterns or characteristics. Current smartphone-based recognition systems depend on the traditional sensors such as accelerometer and gyroscope, which are inbuilt in these devices. Therefore, apart from using information from the traditional sensors, these systems lack contextual information to support automatic activity recognition. In this article, we explore environment contexts such as illumination(light conditions) and noise level to support sensory data obtained from the traditional sensors using a hybrid of Convolutional Neural Networks and Long Short Time Memory(CNN_LSTM) learning models. The models performed sensor fusion by augmenting the low-level sensor signals with rich contextual data to improve recognition and generalisation ability of the proposed solution. Two sets of experiments were performed to validate the proposed solution. The first set of experiments used inertial sensing data whilst the second set of extensive experiments combined inertial signals with contextual information from environment sensing data. Obtained results demonstrate that contextual information such as environment noise level and illumination using hybrid deep learning models achieved better recognition accuracy than the traditional activity recognition models without contextual information.

Nowadays, and with the mechanization of life, speech processing has become so crucial for the interaction between humans and machines. Deep neural networks require a database with enough data for training. The more features are extracted from the speech signal, the more samples are needed to train these networks. Adequate training of these networks can be ensured when there is access to sufficient and varied data in each class. If there is not enough data; it is possible to use data augmentation methods to obtain a database with enough samples. One of the obstacles to developing speech emotion recognition systems is the Data sparsity problem in each class for neural network training. The current study has focused on making a cycle generative adversarial network for data augmentation in a system for speech emotion recognition. For each of the five emotions employed, an adversarial generating network is designed to generate data that is very similar to the main data in that class, as well as differentiate the emotions of the other classes. These networks are taught in an adversarial way to produce feature vectors like each class in the space of the main feature, and then they add to the training sets existing in the database to train the classifier network. Instead of using the common cross-entropy error to train generative adversarial networks and to remove the vanishing gradient problem, Wasserstein Divergence has been used to produce high-quality artificial samples. The suggested network has been tested to be applied for speech emotion recognition using EMODB as training, testing, and evaluating sets, and the quality of artificial data evaluated using two Support Vector Machine (SVM) and Deep Neural Network (DNN) classifiers. Moreover, it has been revealed that extracting and reproducing high-level features from acoustic features, speech emotion recognition with separating five primary emotions has been done with acceptable accuracy.

Nowadays, speed up development and use of digital devices such as smartphones have put people at risk of internet crimes. The evidence of present crimes in a computer file can be easily unreachable by changing the prefix of a file or other algorithms. In more complex cases, either file divided into different parts or the parts of a file that has information about the file type are deleted, where the file fragment recognition issue is discussed. The known files are divided into different fragments, and different classification algorithms to solve the problems of file fragment recognition. A confusion matrix measures the accuracy of type recognition. In the present study, first, the file is divided into different fragments. Then, the file fragment features, which are obtained from Binary Frequency Distribution (BFD), are reduced by 2 feature reduction algorithms; Sequential Forward Selection algorithm (SFS) as well as Sequential Floating Forward Selection algorithm (SFFS) to delete sparse features that result in increased accuracy and speed. Finally, the reduced features are given to 3 classifier algorithms, Multilayer Perceptron (MLP), Support Vector Machines (SVM), and K-Nearest Neighbor (KNN) for classification and comparison of the results. In this paper, we proposed the algorithm of file type recognition that can recognize 6 types of useful files ( pdf, txt, jpg, doc, html, exe), which may distinguish a type of file fragments with higher accuracy than the similar works done.

Over the years, there has been a significant rise in the world's scientific knowledge. However, most of it lacks structure and is often termed as Dark Data. Both humans and expert systems have continually faced difficulty in analyzing and comprehending such overwhelming amounts of information which is crucial in solving several real-world problems. Information and data visualization techniques proffer a promising solution to explore such data by allowing quick comprehension of information, the discovery of emerging trends, identification of relationships and patterns, etc. In this tutorial, we utilize the rich corpus of PubMed comprising of more than 30 million citations from biomedical literature to visually explore and understand the underlying key-insights using various information visualization techniques. With this study, we aim to diminish the limitation of human cognition and perception in handling and examining such large volumes of data by speeding up the process of decision making and pattern recognition and enabling decision-makers to fully understand data insights and make informed decisions.

Studies on crop response to light quality [red (R) to far-red (FR) light ratio] often recommend early weed removal to reduce the effects of shade avoidance responses on crop yield. However, it is unclear whether crops are able to distinguish reflected light quality of kin from that of non-kin. We evaluated the response of sugarbeet (Beta vulgaris L.) to reflected FR light from sugarbeet, common lambsquarters (Chenopodium album L.), Kentucky bluegrass (Poa pratensis L.), alfalfa (Medicago sativa L.), and bare soil (control) under outdoor conditions in 2016 and 2017. Treatments were completely randomized with 10 replications per treatment. The study methods ensured there was no direct resource competition. The reflected R:FR of plant species ranged from 0.06 (common lambsquarters) to 0.24 (sugarbeet) compared to 0.7 for the bare soil. In both 2016 and 2017, there were 2 to 4 more leaves in the sugarbeet surrounded by soil compared to sugarbeet surrounded by neighboring species. There was up to 47, 57, 43, and 23% reduction in sugarbeet leaf area, shoot dry weight, root diameter, and root dry weight, respectively, due to reflected R:FR light from neighboring species. Sugarbeet did not respond differently to reflected light quality of kin compared to non-kin.

Multi-floor environments are usually ignored while designing an autonomous robot for indoor cleaning applications. However, for efficient operation in such environments, the ability of a robotic platform to traverse staircases is crucial. Staircase detection and localization is highly important for planning the traversal on staircases. This paper describes a deep learning approach using Convolutional Neural Networks (CNNs) based Robot Operation System (ROS) to staircase detection and localization. We use an object detection network to detect staircases in images. We also localize these staircases using a contour detection algorithm to detect the target point, a point close to the center of the first step, and the angle of approach to the target point. Experiments are performed with data obtained from images captured on different types of staircases at different view points/angles. Results show that the approach is very accurate in identifying the presence of the staircase in the working environment and is also able to locate the target point with good accuracy.

A supramolecular host-guest interaction and sensing between cationic pillar[5]arenes (CP5) and L-carnitine were developed by the competitive host-guest recognition for the first time. The fluorescence sensing platform was constructed by CP5 functionalized Au nanoparticles (PP5@Au-NPs) as receptor and probe (rhodamine 123, R123), which shown a high sensitivity and selectivity to L-carnitine detection. Due to the property of the negative charge and molecular size of L-carnitine, it can be highly captured by the CP5 via electrostatic interactions and hydrophobic interactions. The mechanism of host-guest between PP5 and L-carnitine was studied by ¹H NMR and molecular docking, which indicated more affinity binding force of PP5 with L-carnitine. Therefore, a selective and sensitive fluorescent method was developed. It has a linear response of 0.1–2.0 and 2.0–25.0 μM and a detection limit of 0.067 μM (S/N = 3) for L-carnitine. The fluorescent sensing platform was also used to detect L-carnitine in human serum and milk samples, which provided potential applications of detection drugs of abuse, and had path for guarding a serious food safety issues.

Exoskeletons designed to assist patients with activities of daily living are becoming increasingly popular, but still are subject to research. In order to gather requirements for the design of such systems, long-term gait observation of the patients over the course of multiple days in an environment of daily living are required. In this paper a wearable all-in-one data acquisition system for collecting and storing biomechanical data in everyday life is proposed. The system is designed to be cost efficient and easy to use, using off-the-shelf components and a cloud server system for centralized data storage. The measurement accuracy of the system was verified, by measuring the angle of the human knee joint at walking speeds between 3 and 12 km/h in reference to an optical motion analysis system. The acquired data was uploaded to a cloud database via a smartphone application. Verification results showed that the proposed toolchain works as desired. The system reached an RMSE from 2.9° to 8°, which is below most of the comparable system. The system provides a powerful, scalable platform for collecting and processing biomechanical data, which can help to automize the generation of an extensive database for human kinematics.

This enables the chip to regulate the voice recognition function of the gadget, leading to a more natural and human-like process. Integrating this technique into the English language voice recognition system can greatly improve its precision in identifying the speech of a particular person. This research seeks to investigate and create an automatic error detection technique for an embedded speech instruction recognition system for English using artificial intelligence. The study commences with a conventional introduction to artificial intelligence, thereafter delving into a comprehensive investigation of the speech recognition algorithm. The MATLAB programme is used to determine the accurate number of words recognised by the system and its accuracy rate. The embedded teaching recognition system for English is then tested in different contexts to assess its performance. Several tests are performed to compare and analyse the results, which indicate that the integrated English voice recognition system achieves an accuracy rate of over 90% and has a low error rate in a calm environment. In a bustling environment with numerous auditory stimuli, the accurate recognition rate typically exceeds 60%.

Visually impaired people encounter many impediments and challenges in their lives such as related to their mobility, education, communication, use of technology, and others. This paper reports the results of an online survey to understand the requirements and challenges blind and visually impaired people face in their daily lives regarding the availability and use of digital devices. The survey was conducted among the blind and visually impaired in Saudi Arabia using digital forms. A total of 164 people responded to the survey most of them using the VoiceOver function. People were asked about the use of smart devices, special devices, operating systems, object recognition apps, indoor and outdoor navigation apps, virtual digital assistive apps, the purpose (navigation, education, etc.) of and difficulty in using these apps, the type of assistance needed, the reliance on others in using the assistive technologies, and the level of satisfaction from the existing assistive technologies. The majority of the participants were 18 – 65 years old with 13% under 18 and 3% above 65. Sixty-five percent of the participants were graduates or postgraduates and the rest only had secondary education. White Cane, mobile phones, Apple iOS, Envision, Seeing AI, VoiceOver, and Google Maps were the most used devices, technologies, and apps used by the participants. Navigation at 39.6% was the most reported purpose of the special devices followed by education (34.1%) and office jobs (12.8%). The information from this survey along with a detailed literature review of academic and commercial technologies for the visually impaired was used to establish the research gap, design requirements, and a comprehensive understanding of the relevant landscape, which in turn was used to design smart glasses called LidSonic for visually impaired.

At present, indoor localization is one of the core technologies of location-based services (LBS), and there exist numerous scenario-oriented application solutions. Visual features, as the main semantic information to help people understand the environment and thus occupy the dominant part, many techniques about indoor scene recognition are widely adopted. However, the engineering application problem of cell phone indoor scene recognition and localization has not been well solved due to insufficient semantic constraint information of building map and the immaturity of building map location anchors (MLA) matching positioning technology. To address the above problems, this paper proposes a cell phone indoor scene recognition and localization method with building map semantic constraints. Firstly, we build a library of geocoded entities for building map location anchors (MLA), which can provide users with "immersive" real-world building maps on the one hand and semantic anchor point constraints for cell phone positioning on the other. Secondly, using the improved YOLOv5s deep learning model carried on the mobile terminal, we recognize the universal map location anchors (MLA) elements in building scenes by cell phone camera video in real-time. Lastly, the spatial location of the scene elements obtained from the cell phone video recognition is matched with the building MLA to achieve real-time positioning and navigation. The experimental results show that the model recognition accuracy of this method is above 97.2%, and the maximum localization error is within the range of 0.775 m, and minimized to 0.5 m after applying the BIMPN road network walking node constraint, which can effectively achieve high positioning accuracy in the building scenes with rich MLA element information. In addition, the building map location anchors (MLA) has universal characteristics, and the positioning algorithm based on scene element recognition is compatible with the extension of indoor map data types, so this method has good prospects for engineering applications.

Facial expression recognition is crucial for understanding human emotions and nonverbal communication. With the growing prevalence of facial recognition technology and its various applications, accurate and efficient facial expression recognition has become a significant research area. However, most previous methods have focused on designing unique deep-learning architectures while overlooking the loss function. This study presents a new loss function that allows simultaneous consideration of inter- and intra-class variations to be applied to CNN architecture for facial expression recognition. More concretely, this loss function reduces the intra-class variations by minimizing the distances between the deep features and their corresponding class centers. It also increases the inter-class variations by maximizing the distances between deep features and their non-corresponding class centers, and the distances between different class centers. Numerical results from several benchmark facial expression databases, such as Cohn-Kanade Plus, Oulu-Casia, MMI, and FER2013, are provided to prove the capability of the proposed loss function compared with existing ones.

Designing motion representations for the problem of 3D human action recognition from skeleton sequences is an important yet challenging task. An effective representation should be robust to noise, invariant to viewpoint changes and result in a good performance with low-computational demand. Two main challenges in this task include how to efficiently represent spatio-temporal patterns of skeletal movements and how to learn their discriminative features for classification task. This paper presents a novel skeleton-based representation and a deep learning framework for 3D action recognition using RGB-D sensors. We propose to build an action map called SPMF (Skeleton Posture-Motion Feature), which is a compact image representation built from skeleton poses and their motions. An Adaptive Histogram Equalization (AHE) algorithm is then applied on the SPMF to enhance their local patterns and form an enhanced action map, namely Enhanced-SPMF. For learning and classification tasks, we exploit Deep Convolutional Neural Networks based on the DenseNet architecture to learn directly an end-to-end mapping between input skeleton sequences and their action labels via the Enhanced-SPMFs. The proposed method is evaluated on four challenging benchmark datasets, including both individual actions, interactions, multiview and large-scale datasets. The experimental results demonstrate that the proposed method outperforms previous state-of-the-art approaches on all benchmark tasks, whilst requiring low computational time for training and inference.

A multitude of factors influence our well-being and health in everyday life. Some of which we are totally aware like a delayed train, car exhausts, or continuous work stress. Other factors influence us on a subconscious level, making them, however, not less important to address. One of these factors is the built environment surrounding us almost every second of our lives. The idea of biophilic design states the importance of natural elements implemented in architectural structures to improve the occupants’ health and well-being. Adding to this field of research, this article targets the impact of biophilic design on positive emotions and productivity by using novel approaches like Facial Emotion Recognition (FER) with Residual Masking Networks, and sentiment detection using Large Language Models. Two separate but related studies were conducted and their results reported here. The first study measured emotions of people when confronted with images of different kinds of architecture, via FER (Facial Emotion Recognition) and also via a user survey. We distinguish between self-stated emotions entered through a survey by study participants, and emotions detected from studying the participants’ faces through AI from our FER (Facial Emotion Recognition) system. We found clear trends for most detected emotions, and significant evidence for self-stated emotions, that architecture implementing biophilic design evokes more positive emotions than one that does not. The second study measures the influence of natural elements on productivity and team-engagement. We found that natural elements do influence productivity and sentiment positively; however, only the change in sentiment was shown to be significant. As our sample size, especially for the second study, was relatively small, future research will apply these ideas in a larger setup to acquire further evidence for the importance of biophilic design for human well-being and health.

Intrinsically disordered proteins mediate crucial biological functions through their interactions with other proteins. Mutual synergistic folding (MSF) occurs when all interacting proteins are disordered, folding into a stable structure in the course of the complex formation. In these cases, the folding and binding processes occur in parallel, lending the resulting structures uniquely heterogeneous features. Currently there are no dedicated classification approaches that would take into account the particular biological and biophysical properties of MSF complexes. Here we present a scalable clustering-based classification scheme, built on redundancy-filtered features that describe the sequence and structure properties of the complexes, and the role of the interaction, which is directly responsible for structure formation. Using this approach, we define six major types of MSF complexes, corresponding to biologically meaningful groups. Hence, the presented method also shows that differences in binding strength, subcellular localization, and regulation are encoded in the sequence and structural properties of proteins. While current structure classification methods can also handle complex structures, we show that the developed scheme is fundamentally different, and since it takes into account defining features of MSF complexes, it serves as a better representation of structures arising through this specific interaction mode.

Automatic recognition methods for non-stationary EEG data collected from EEG sensors play an essential role in neurological detection. The integrative approaches proposed in this study consists of Symlet wavelet processing, a gradient boosting machine, and a grid search optimizer for a three-level classification scheme for normal subjects, intermittent epilepsy, and continuous epilepsy. Fourth-order Symlet wavelets were adopted to decompose the EEG data into five time-frequency sub-bands, whose statistical features were computed and used as classification features. The grid search optimizer was used to automatically find the optimal parameters for training the classifier. The classification accuracy of the gradient boosting machine was compared with that of a support vector machine and a random forest classifier constructed according to previous descriptions. Multiple-index were used to evaluate the Symlet wavelet transform-gradient boosting machine-grid search optimizer classification scheme, which provided better classification accuracy and detection effectiveness than has recently reported in other work on three-level classification of EEG data.

We introduce the Open Research Knowledge Graph Agriculture Named Entity Recognition (the ORKG Agri-NER) corpus and service for contribution-centric scientific entity extraction and classification. The ORKG Agri-NER corpus is a seminal benchmark for the evaluation of contribution-centric scientific entity extraction and classification in the agricultural domain. It comprises titles of scholarly papers that are available as Open Access articles on a major publishing platform. We describe the creation of this corpus and highlight the obtained findings in terms of the following features: 1) a generic conceptual formalism focused on capturing scientific entities in agriculture that reflect the direct contribution of a work; 2) a performance benchmark for named entity recognition of scientific entities in the agricultural domain by empirically evaluating various state-of-the-art sequence labeling neural architectures and transformer models; and 3) a delineated 3-step automatic entity resolution procedure for the resolution of the scientific entities to an authoritative ontology, specifically AGROVOC that is released in the Linked Open Vocabularies cloud. With this work we aim to provide a strong foundation for future work on the automatic discovery of scientific entities in the scholarly literature of the agricultural domain.

In the dynamic field of mental health care, the nuanced application of Artificial Intelligence (AI) through Natural Language Processing (NLP) and Automatic Speech Recognition (ASR) is a pivotal innovation in patient empowerment and service optimization. This study introduces a distinctive online therapy platform that capitalizes on the synergy of NLP and ASR to offer unprecedented levels of interactive and personalized therapeutic interventions. The architecture of our system is meticulously detailed, featuring an ASR component with an impressive Word Error Rate (WER) of 14% when trained on the diverse French subsets of the Mozilla Common Voice dataset, complemented by a high-precision NLP framework skilled in processing and responding to user inputs. The evaluation of our system highlights its efficacy in enhancing therapy sessions and user satisfaction, with an emphasis on the qualitative aspects of user feedback. The paper addresses challenges such as dataset representativeness and language model refinement and articulates the strategic solutions employed to overcome them. The paper concludes with forward-looking perspectives on AI's role in mental health services, advocating for the creation of sophisticated, language-specific datasets and models to satisfy the increasing demands of online therapy, reflecting a growing commitment of patients in the management of their therapy. This research underscores the transformative impact of AI in advancing mental health care into the digital age, representing a significant evolution over existing methodologies.

Human violence recognition is an area of great interest in the scientific community, given its broad spectrum of applications, especially in video surveillance systems, since detecting violence in real-time could prevent criminal acts and save lives. Despite the number of existing proposals and research, most focus on the precision of results, leaving aside efficiency and its practical implementation. Thus, this work proposes a model that is effective and efficient in recognizing human violence in real-time. The proposed model consists of three modules: a first module called Spatial Motion Extractor (SME), in charge of extracting regions of interest from a frame; a second module called Short Temporal Extractor (STC), whose function is to extract temporal characteristics of rapid movements, finally the Global Temporal Extractor (GET) module, responsible for identifying long-lasting temporal features and fine-tuning the model. The proposal was evaluated regarding efficiency, effectiveness, and ability to operate in real-time. The results obtained on Hockey, Movies, and RWF-2000 datasets demonstrated that this approach is highly efficient compared to other alternatives. A VioPeru dataset was created to validate real-time applicability with violent and non-violent videos captured by real video surveillance cameras in Peru. The effectiveness results in this dataset outperformed the best existing proposal. Therefore, our proposal has contributions in efficiency, effectiveness, and real-time.

Emotions play a vital role in understanding human behavior and interpersonal relationships. The ability to recognize emotions through Electroencephalogram (EEG) signals offers an alternative to traditional methods, such as questionnaires, enabling the identification of emotional states in a non-intrusive manner. Automatic emotion recognition holds great potential, eliminating the need for clinical examinations or physical visits, thereby contributing significantly to the advancement of Brain-Computer Interface (BCI) technology. However, one of the key challenges lies in effectively selecting and extracting relevant features from the EEG signal to establish meaningful distinctions between different emotional states. The process of feature selection is often time-consuming and demanding. In this research, we propose a groundbreaking approach for automatically identifying three emotional states (positive, negative, and neutral) by leveraging auditory stimulation of EEG signals. Our novel method directly applies the raw EEG signal to a Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) architecture, bypassing the conventional feature extraction and selection steps. This unconventional approach offers a significant departure from existing literature. Our proposed network architecture comprises ten convolutional layers, followed by three LSTM layers and two fully connected layers. Through extensive simulations and evaluations on 12 active channels, our algorithm demonstrates exceptional performance, achieving an accuracy of 97.42\% and 95.23\% for the binary classification of negative and positive emotions, as well as a Cohen's Kappa coefficient of 0.96 and 0.93 for the three-class classification (negative, neutral, and positive), respectively. These promising results highlight the efficacy of our novel methodology and its potential implications in advancing emotion recognition using EEG signals.

The progressive impairment analysis in gait from neurological diseases patients such as Hereditary Ataxias (HA) has been carried out using gait data collected with movement sensors. This research is focused on finding the minimum amount required of gait features to recognize efficiently and less intrusive way, HA patients based on data collected with iPhone movement sensors placed on the ankles from 14 HA patients and 14 healthy people. A twofold proposal is made , first a local minimum prominent peak criterion to find out the starting point of each stride, to get 10-stride window about which 56 spatial-temporal features are derived; second a search strategy based on Hill Climbing algorithm to reduce the number of gait features and sensors. The main results were the findings that with two gait patterns a 96% of classification accuracy was achieved by using K-Nearest Neighbors (KNN) and Multi-Layer Perceptron (MLP) algorithms, but in addition, MLP only right ankle sensor patterns were required which also allows to reduce the intrusion.

Previous studies to recognize negative emotions (e.g. disgust, fear, sadness) for mental health care have used heavy equipment directly attaching electroencephalogram (EEG) electrodes to the head, making it difficult to use in daily life, and they have proposed binary classification methods to determine whether negative emotion or not. To tackle this problem, we propose a negative emotion recognition system to collect multimodal biosignal data such as five EEG signals in an EEG headset and heart rate, galvanic skin response, and skin temperature in a smart band for classifying multiple negative emotions. It consists of android Internet of Things (IoT) application, an oneM2M-compliant IoT server, and a machine learning server. The android IoT application upload the biosignal data to the IoT server. By using the biosignal data stored in the IoT server, the machine learning server recognizes the negative emotions of disgust, fear, and sadness using a multi-class support vector machine (SVM) model with a radial basis function kernel (RBF). The experimental results showed that the multi-class SVM model achieved 93% accuracy when considering all the multimodal biosignal data. Moreover, when considering only data in the smart band, it could achieve 98% accuracy by optimizing the hyper-parameter of the RBF kernel.

Speech emotion recognition is a critical component for achieving natural human-robot interaction. The modulation-filtered cochleagram is a feature based on auditory modulation perception, which contains multi-dimensional spectral-temporal modulation representation. In this study, we propose an emotion recognition framework that utilizes a multi-level attention network to recognize emotions from the modulation-filtered cochleagram. The channel-level attention and spatial-level attention modules are used to capture emotional saliency maps of channel and spatial feature representations from the 3D convolution feature maps, respectively. Furthermore, the temporal-level attention module captures significant emotional regions from the concatenated feature sequence of the emotional saliency maps. Our experiments on the IEMOCAP dataset demonstrate that the modulation-filtered cochleagram significantly improves the prediction performance of categorical emotion compared to other evaluated features. Moreover, our emotion recognition framework achieves a better unweighted accuracy of 71% in categorical emotion recognition than several existing approaches in the experiments. In summary, our study demonstrates the effectiveness of the modulation-filtered cochleagram in speech emotion recognition, and our proposed multi-level attention framework provides a promising direction for future research in this field.

In the context of Shared Autonomous Vehicles, the need to monitor the environment inside the car will be crucial. This article focuses on the application of deep learning algorithms to detect objects, namely lost/forgotten items to inform the passengers, and aggressive items to monitoring if violent actions may arise between passengers. For object detection algorithms was used public datasets (COCO and TAO) to train state-of-the-art algorithms, such as YOLOv5. For violent action detection was used the MoLa InCar dataset to train on state-of-the-art algorithms such as I3D, R(2+1)D, SlowFast, TSN and TSM. At the end an embedded automotive solution was used to demonstrate both methods running in real-time.

In the current manuscript, a novel software application for Rapid Damage Assessment of RC buildings subjected to earthquake excitation is presented based on Artificial Neural Networks. The software integrates the use of a novel deep learning methodology for Rapid Damage Assessment into modern software development platforms, while the developed graphical user interface promotes the ease of use even from non-experts. The aim is to foster actions both in the pre- and post- earthquake phase. The structure of the source code permits the usage of the application either autonomously as a software tool for Rapid Visual Inspections of buildings prior to or after a strong seismic event or as a component of Building Information Modelling systems in the framework of digitizing building data and properties. The methodology implemented for the estimation of the RC buildings’ damage states is based on the theory and algorithms of Pattern Recognition problems. The effectiveness of the developed software is successfully tested using an extended, numerically generated database of RC buildings subjected to recorded seismic events

The 21st century, where all things are depending upon technology, almost all the human tasks are being done with the help of technology to save a lot of time and make our life much more comfortable. The monitoring of students’ attendance is a crucial task for faculty in today’s world. There are more chances of errors while entering the students’ attendance records in the primary system, primarily when the class was over. The main concern of this study is to build an IoT-based automated attendance management system for educational institutes by biometric recognition to incorporate fake/proxy attendance and errors of entry and to replace old manual methods of taking students’ attendance by calling their names or roll numbers. The AAS will click the image of the classroom, and it will automatically detect the faces of students sitting in the lecture room and recognize them during lectures then mark their attendance daily to keep a record of their presence and also maintain and manage it for the management staff of the institution for future by using web services.

Cardiovascular disease (CVD) may sometimes unexpected loss of life. It affects the heart and blood vessels of body. CVD plays an important factor of life since it may cause death of human. It is necessary to detect early of this disease for securing patients life. In this chpter two exclusively different methods are proposed for detection of heart disease. The first one is Pattern Recognition Approach with grammatical concept and the second one is machine learning approach. In the syntactic pattern recognition approach initially ECG wave from different leads is decomposed into pattern primitive based on diagnostic criteria. These primitives are then used as terminals of the proposed grammar. Pattern primitives are then input to the grammar. The parsing table is created in a tabular form. It finally indicates the patient with any disease or normal. Here five diseases beside normal are considered. Different Machine Learning (ML) approaches may be used for detecting patients with CVD and assisting health care systems also. These are useful for learning and utilizing the patterns discovered from large databases. It applies to a set of information in order to recognize underlying relationship patterns from the information set. It is basically a learning stage. Unknown incoming set of patterns can be tested using these methods. Due to its self-adaptive structure Deep Learning (DL) can process information with minimal processing time. DL exemplifies the use of neural network. A predictive model follows DL techniques for analyzing and assessing patients with heart disease. A hybrid approach based on Convolutional Layer and Gated-Recurrent Unit (GRU) are used in the paper for diagnosing the heart disease.

With the evolution of computing technology in many application like human robot interaction, human computer interaction and health-care system, 3D human body models and their dynamic motions has gained popularity. Human performance accompanies human body shapes and their relative motions. Research on human activity recognition is structured around how the complex movement of a human body is identified and analyzed. Vision based action recognition from video is such kind of tasks where actions are inferred by observing the complete set of action sequence performed by human. Many techniques have been revised over the recent decades in order to develop a robust as well as effective framework for action recognition. In this survey, we summarize recent advances in human action recognition, namely the machine learning approach, deep learning approach and evaluation of these approaches.

The identification of bloodstain is one of the most important approaches in obtaining evidence in criminalistics. A threshold method based on spectral co-efficient and interclass variance is proposed in this paper, it is a non-contact, non-destructive method for quickly identifying bloodstains. The spectra of bloodstains and other suspected substances were all extracted from their hyper-spectral image. Then calculate the correlation coefficients of these spectral and interclass variances, analyze the differences between substances. The best blood recognition threshold was determined as 0.9. After preprocessing for eliminating systematic errors, experiments with the threshold 0.9 are carried out to identify bloodstains on the calico and red T-shirt. The method can remarkably identify the bloodstain from other non-blood substances both quickly and efficiently. The blood extraction rate can reach to 93.35% and 89.19%, respectively. It is an important step toward the implementation of bloodstain non-contact and non-destructive identification in forensic casework.

Nowadays, the combination of fingerprinting methodology with friendly environmental and economical analytical instrumentation are becoming increasingly relevant in the food sector. In this study, a highly versatile portable analyser based on Spatially Offset Raman Spectroscopy (SORS) to obtain the edible vegetable oils (sunflower and olive oils) fingerprints was used to evaluate the capability of such fingerprints, obtained quickly, reliable and without any sample treatment, to discriminate/classify the analysed samples. After data treatment, not only HCA and PCA as unsupervised pattern recognition techniques but also SVM, kNN and SIMCA as supervised pattern recognition techniques, showed that the main effect over the discrimination/classification was associated to those regions of RAMAN fingerprint related to the free fatty acids content, especially oleic and linoleic acid. These facts allowed the discrimination attending to the original raw material used in the oil's elaboration. In all the model established, reliable qualimetric parameters were obtained.

Fermatean fuzzy sets (FFSs) are a versatile tool for handling uncertain problems and have shown great effectiveness in practical applications. However, many existing Fermatean fuzzy similarity measures exhibit counter-intuitive situations, making it challenging to accurately measure the similarity or difference between FFSs. To address this issue, several similarity and distance measures for FFSs are proposed, inspired by the Tanimoto similarity measure. The properties of the proposed measures are also explored, along with several comparative examples with existing measures for FFSs, which illustrate their superior performance in processing fuzzy information from FFSs. The proposed measures have effectively overcome the counter-intuitive challenges posed by many existing measures and significantly outperforms existing measures in differentiating different FFSs. Furthermore, we demonstrate the practical applicability of the proposed measures in solving pattern recognition, medical diagnosis and multiple-attribute decision-making problems.

Theories of reading comprehension emphasize decoding and listening comprehension as two essential components. The current study aimed to investigate how Chinese character decoding and context-driven auditory semantic integration contribute to reading comprehension in Chinese middle school students. Seventy-five middle school students were tested. Context-driven auditory semantic integration was assessed with speech-in-noise tests in which the fundamental frequency (F0) contours of spoken sentences were either kept natural or acoustically flattened with the latter requiring a higher degree of contextual information. Statistical modelling with hierarchical regression was conducted to examine the contributions of Chinese character decoding and context-driven auditory semantic integration to reading comprehension. Performance on Chinese character decoding and auditory semantic integration scores with the flattened (but not natural) F0 sentences significantly predicted reading comprehension. Furthermore, the contributions of these two factors to reading comprehension were better fitted with an additive model instead of a multiplicative model. These findings indicate that reading comprehension in middle schoolers is associated with not only character decoding but also the listening ability to make better use of the sentential context for semantic integration in a severely degraded speech-in-noise condition. The results add to our better understanding of the multi-faceted reading comprehension in children. Future research could further address the age-dependent development and maturation of reading skills by examining and controlling other important cognitive variables, and apply neuroimaging techniques such as functional magmatic resonance imaging to reveal the neural substrates for the contribution of auditory semantic integration and the observed additive model to reading comprehension.

Considering Cu(tda) chelate is a receptor for adenine and related 6-aminopurines, now we report the synthesis, molecular and crystal structures, thermal stability, spectral properties and DFT calculations on the [Cu(tda)(9heade)(H2O)]•2H2O (1). N7 is found to be the preferred donor atom of 9heade, that can favor the cooperation of a Cu-N7(9heade) bond and an (9heade)N6-H•••O(tda) interligand interaction. But unprecedentedly, the Cu(tda)-9heade molecular recognition in 1 is essentially featured the Cu-N1(9heade), without any N6-H•••O(carboxyl tda) interligand interaction. Nevertheless, being N1 the most basic donor for N9-substituted adenines, this Cu-N1 bond is now assisted by an O2-water mediated interaction (N6-H•••O2 and O2•••Cu weak contact). Also in the crystal packing, the O-H(ol) of 9heade interacts with its own adenine moiety by an O3-water mediated interaction (O(ol)-H•••O3 plus O3-H36•••π(adenine moiety). In turn these water-mediated interactions seems to be responsible to serious alterations on the physical properties of crystalline or grounded samples. Density functional theory calculations were used to evaluate the interactions energetically. Moreover, the quantum theory of atoms-in-molecules (QTAIM) in combination with the noncovalent interaction plot (NCIPlot) were used to analyze the interactions and rationalize the existence and relative importance of hydrogen bonding, chalcogen bonding and π-stacking interactions.

Wernicke-Korsakoff syndrome (WKS) is induced by thiamine deficiency (TD) and mainly related to alcohol consumption. Frontal cortex dysfunction has been associated to impulsivity and disinhibition in WKS patients. The pathophysiology involves oxidative stress, excitotoxicity and inflammatory responses leading to neuronal death, but the relative contributions of each factor (alcohol and TD, isolate or in interaction) to these phenomena are still poorly understood. A rat model was used by forced consumption of 20% (w/v) alcohol for 9 months (CA), TD hit (TD diet + pyrithiamine 0.25 mg/kg, i.p. daily injections the last 12 days of experimentation; TDD), and both combined treatments (CA+TDD). Motor and cognitive performance and cortical damage were examined. CA caused hyperlocomotion as a possible sensitization of ethanol-induced excitatory effects and recognition memory deficits. In addition, CA+TDD animals showed a disinhibited-like behavior, which appears to be dependent on TDD. Also, combined treatment led to more pronounced alterations in nitrosative stress, lipid peroxidation, apoptosis and cell damage markers. Correlations between injury signals and disinhibition suggest that CA+TDD disrupts behaviors dependent on the frontal cortex. Our study sheds light on the potential disease-specific mechanisms, reinforcing the need for neuroprotective therapeutic approaches along with preventive treatments for the nutritional deficiency in WKS.

Accurate and robust deep learning models for facial expression recognition are challenging to achieve, given the diversity of human faces and variations in images, including different facial poses and lighting conditions. In this work, we proposed a clock-driven convolutional Spiking Neural Network (SNN) and STDP fine-tune architecture, meticulously calibrated its hyperpa-rameters, and experimented with various optimization methods. The best model resulted was trained and evaluated on the Fer2013 and FER+ database, obtaining an accuracy of 61.87% and 79.97% without requiring auxiliary training data or face registration. To our best knowledge, the proposed SNN achieved comparable accuracy to CNNs of similar depth and possessed advantages of low energy consumption and high computational efficiency. The computational efficiency of the proposed SNN is approximately three times that of CNNs. Along with this, we introduced the very recent cumulative spike guided encoder visualization technique and revealed the strong encoding capability of the proposed SNN.

Machine learning is a fundamental aspect of artificial intelligence that involves the development of algorithms and models that allow computers to learn and make predictions or decisions without explicit programming. With the development of neural networks, back-propagation algorithms and deep learning, machine learning has made breakthroughs in the fields of image recognition, natural language processing and handwriting recognition using machine learning techniques. The advent of deep learning has revolutionized the field of handwriting recognition, using convolutional neural networks, recurrent neural networks, and sequence-to-sequence models to provide solutions that go beyond machine learning methods and significantly improve the accuracy and robustness of handwriting recognition systems. But challenges remain, including the need for large labelled datasets, computational resources and addressing potential biases. As research in deep learning techniques continues to drive handwriting recognition closer towards realisability, machine learning approaches remain at the forefront.

Autonomous target recognition (ATR) plays a crucial role in maintaining situational awareness during environmental monitoring. Unmanned aerial vehicles (UAVs) equipped with autonomous target recognition technology can gather and analyze real-time information about targets, including their locations, sizes, and types. However, UAV-captured images in complex real-world environments often display significant variations in perspective and scale due to changes in UAV altitude and distance. Existing methods for autonomous target recognition on UAVs struggle to capture targets from large field-of-view and multi-scale images, resulting in low recognition accuracy and high false-positive rates. This paper introduces two novel Slimmable neural network models, namely SE-YOLOv5s and ST-YOLOv5s, which are based on the YOLOv5s architecture. These models incorporate the Squeeze and Excitation and Swin-Transformer mechanisms to enhance the ability to extract features from large field-of-view and multi-scale images. To evaluate their performance, experiments were conducted on the Visdrone19 aerial dataset. Compared to the state-of-the-art YOLOv5s algorithm, the utilization of SE-YOLOv5s and ST-YOLOv5s for autonomous target recognition on low-altitude drones resulted in improvements in both accuracy and false-positive rates. These proposed methods combine Slimmable neural network design with feature enhancement mechanisms, addressing the challenges posed by complex real-world environments in UAV missions. The advancements in autonomous target recognition on low-altitude drones will significantly contribute to enhancing situational awareness in future environmental monitoring.

Accurate prediction of the degree of airport delays under the influence of convective weather is crucial for collaborative traffic management implementation and improving the efficiency of airport operations. However, existing studies usually only consider numerical-type quantitative features of weather-affected traffic in their models, and lack the introduction of spatial information to comprehensively portray the traffic operation scenarios under the influence of weather. In order to overcome this problem, this paper firstly designs a new image representation of weather-affected air traffic, and constructs a multi-channel traffic and weather scene image (MTWSI) by populating the airspace two-dimensional grid with traffic and meteorological information to represent the overall traffic operation situation in the terminal area under the influence of convective weather; then, a deep convolutional neural network-based airport delay prediction model ( ADLCNN), which takes MTWSI images as input and uses a specific CNN model to extract the deep features that affect traffic operation in it to input into the subsequent classification algorithm to predict the flight delay level; finally, a series of comparative experiments are carried out on the operational data of Guangzhou Baiyun Airport, and the experimental validation shows that, compared with the traditional machine learning methods, the proposed CNN-based airport delay prediction The experimental validation shows that the proposed CNN-based airport delay prediction model has satisfactory prediction performance compared with the traditional machine learning methods, which also proves that the proposed MTWSI method can more comprehensively respond to the real traffic conditions.

Human activity recognition (HAR) using wearable sensors is gaining popularity in the healthcare domain, enabling continuous monitoring and timely interventions. The conventional HAR systems are centralised and need data sharing with a server for processing and model training. However, this data sharing leads to various challenges, such as privacy concerns, increased communication and storage costs, inefficient computational algorithms, data scarcity due to unwillingness to share, and difficulty achieving high accuracy and personalised models simultaneously. To overcome these challenges, this paper introduces a privacy and energy-aware federated learning (FL) framework for model training without data sharing. The proposed scheme incorporates spiking neural networks (SNNs), which are particularly suited for resource-constrained devices due to their event-driven information processing. The key idea is to synergise the strengths of SNNs and long short-term memory (LSTM) to create a hybrid model for robust HAR. The proposed hybrid spiking LSTM (S-LSTM) model uses the LSTM as an input layer to capture temporal dependencies efficiently, followed by spiking layers that replace the typical activation using leaky integrate-and-fire neurons. We employ the combination of surrogate gradient learning and backpropagation through time (BPTT) to update the weights of SNN layers in a supervised manner. The performance of the proposed model is evaluated using two publicly available datasets for both indoor and outdoor scenarios. The results show that the proposed S-LSTM performs much better than spike convolutional neural networks (S-CNN) and simple CNN, achieving the accuracy of 97% and 87% for indoor and outdoor datasets, respectively. Personalisation, also known as fine-tuning the global model using local data, is crucial in HAR. It allows the model to adapt to individual behaviour and preferences, making predictions more accurate. The results also show that personalisation improves the performance of participants up to 9% on average. Additionally, analysis with 50% random client participation reduces communication costs by 50% with minimal impact on accuracy.

Capsule endoscopy (CE) has been a widely used medical imaging tool for the diagnosis of gastrointestinal tract abnormalities like bleeding. But the CE captures a huge number of image frames that are time-consuming and tedious tasks for medical experts to diagnose manually. To address this issue, researchers focused on the computer-aided bleeding detection system to identify bleeding automatically in real-time. This paper presents a systematic review of the available state-of-the-art computer-aided bleeding detection algorithms for capsule endoscopy. The review was carried out by searching five different repositories: Scopus, PubMed, IEEE Xplore, ACM Digital Library, and ScienceDirect for all original publications on computer-aided bleeding detection published between 2001 and 2021. The PRISMA methodology was used to perform the review, and 112 full-texts of scientific papers were reviewed. The contributions of this paper are I) a taxonomy for computer-aided bleeding detection algorithms for capsule endoscopy is identified, II) the available state-of-the-art computer-aided bleeding detection algorithms including various color spaces (RGB, HSV, etc.), feature extraction techniques, and classifiers are discussed, and III) identify the most effective algorithm for practical use cases. Finally, the paper is concluded by providing future direction for computer-aided bleeding detection research.

The classifier system proposed in this work combines the dissimilarity spaces produced by a set of Siamese neural networks (SNNs) designed using 4 different backbones, with different clustering techniques for training SVMs for automated animal audio classification. The system is evaluated on two animal audio datasets: one for cat and another for bird vocalizations. Different clustering methods reduce the spectrograms in the dataset to a set of centroids that generate (in both a supervised and unsupervised fashion) the dissimilarity space through the Siamese networks. In addition to feeding the SNNs with spectrograms, additional experiments process the spectrograms using the Heterogeneous Auto-Similarities of Characteristics. Once the similarity spaces are computed, a vector space representation of each pattern is generated that is then trained on a Support Vector Machine (SVM) to classify a spectrogram by its dissimilarity vector. Results demonstrate that the proposed approach performs competitively (without ad-hoc optimization of the clustering methods) on both animal vocalization datasets. To further demonstrate the power of the proposed system, the best stand-alone approach is also evaluated on the challenging Dataset for Environmental Sound Classification (ESC50) dataset. The MATLAB code used in this study is available at https://github.com/LorisNanni.

Research on Traffic Light Recognition (TLR) has grown in recent years, primarily driven by the growing interest in autonomous vehicles development. Machine Learning algorithms have been widely used to that purpose. Mainstream approaches, however, require large amount of data to properly work, and as a consequence, a lot of computational resources. In this paper we propose the use of Expert Instruction (IE) as a mechanism to reduce the amount of data required to provide accurate ML models for TLR. Given an image of the exterior scene taken from the inside of the vehicle, we stand the hypothesis that the picture of a traffic light is more likely to appear in the central and upper regions of the image. Frequency Maps of traffic light location were thus constructed to confirm this hypothesis. The frequency maps are the result of a manual effort of human experts in annotating each image with the coordinates of the region where the traffic light appears. Results show that EI increased the accuracy obtained by the classification algorithm in two different image datasets by at least 15%. Evaluation rates achieved by the inclusion of EI were also higher in further experiments, including traffic light detection followed by classification by the trained algorithm. The inclusion of EI in the PCANet achieved a precision of 83% and recall of 73% against 75.3% and 51.1%, respectively, of its counterpart. We finally presents a prototype of a TLR Device with that expert model embedded to assist drivers. The TLR uses a smartphone as a camera and processing unit. To show the feasibility of the apparatus, a dataset was obtained in real time usage and tested in an Adaptive Background Suppression Filter (AdaBSF) and Support Vector Machines (SVMs) algorithm to detect and recognize traffic lights. Results show precision of 100% and recall of 65%.

In modern security situations, tracking multiple human objects in real-time within challenging urban environments is a critical capability for enhancing situational awareness, minimizing response time, and increasing overall operational effectiveness. Tracking multiple entities enables informed decision-making, risk mitigation, and the safeguarding of civil-military operations to ensure safety and mission success. This paper presents a multi-modal electro-optical/infrared (EO/IR) and radio frequency (RF) fused sensing (MEIRFS) platform for real-time human object detection, recognition, classification, and tracking in challenging environments. By utilizing different sensors in a complementary manner, the robustness of the sensing system is enhanced, enabling reliable detection and recognition results across various situations. Specifically designed Radar tag and thermal tag can be used to discriminate friendly and non-friendly objects. The system incorporates deep learning-based image fusion and human object recognition and tracking (HORT) algorithms to ensure accurate situation assessment. After integrating into an all-terrain robot, multiple ground tests were conducted to verify the consistent HORT in various environments. The MEIRFS sensor system has been designed to meet the Size, Weight, Power, and Cost (SWaP-C) requirements for installation on autonomous ground and aerial vehicles.

Speech emotion recognition (SER) is a technology that can be applied in distance education to analyze speech patterns and evaluate speakers’ emotional states in real-time. It provides valuable insights and can be used to enhance the learning experience by enabling the assessment of instructors’ emotional stability, a factor that significantly impacts information delivery effectiveness. Students demonstrate different engagement levels during learning activities, and assessing this engagement is an important aspect of controlling the learning process and improving e-learning systems. An important aspect that may influence student engagement is the emotional states of their instructors. Accordingly, this research uses deep learning techniques to create an automated system for recognizing instructors’ emotions in their speech when delivering distance learning. This methodology entails integrating Transformer, convolutional neural network, and long short-term memory architectures into an ensemble to enhance SER. Feature extraction from audio data used Mel-frequency cepstral coefficients, chroma, Mel spectrogram, zero crossing rate, spectral contrast, centroid, bandwidth, roll-off, and root-mean square, with subsequent optimization processes adding noise to, conducting time stretching, and shifting the audio data. Notably, several Transformer blocks were incorporated, and a multi-head self-attention mechanism was employed to identify the relationships between the input sequence segments. The pre-processing and data augmentation methodologies significantly enhanced the precision of the results in that the model achieved accuracy rates of 96.3%, 99.86%, 96.5%, and 85.3% on the Ryerson Audio-Visual Database of Emotional Speech and Song, Berlin Database of Emotional Speech, Surrey Audio-Visual Expressed Emotion, and Interactive Emotional Dyadic Motion Capture datasets. Furthermore, it achieved 83% accuracy on another dataset created for this research—the Saudi Higher Education Instructor Emotions dataset. The results demonstrate this model’s considerable accuracy in detecting emotions in speech data across different languages and datasets.

One of the most important component of a Geographic Information Retrieval (GIR) is the geoparser, which performs toponym recognition, disambiguation, and geographic coordinate resolution from unstructured text domain. However, news articles which report several events across many place references mentioned in the document is not yet adequately modeled by regular geoparser types where the scope of resolution is either on toponym-level or document-level. The capacity to detect multiple events, geolocate its true locations and coordinates along with their numerical arguments are still missing from modern geoparsers, much less in Indonesian news corpora domain. We propose a novel type event geoparser which integrates an ACE-based event extraction model and provides precise event-level scope resolution. The geoparser casts the geotagging and event extraction as sequence labeling and uses Conditional Random Field with keywords feature obtained using Aggregated Topic Model as a semantic exploration from large corpus, which eventually increases the generalizability of the model. The geoparser also use Smallest Administrative Level feature along with Spatial Minimality-derived algorithm to improve the identification of Pseudo-location entities, resulting 19.4% increase for weighted F1 score. As a side effect of event extraction, the geoparser also extracts various numerical arguments and able to generate thematic choropleth map from a single news story.

The integration of deep learning methods, especially Convolutional Neural Networks (CNN), and Synthetic Aperture Radar Automatic Target Recognition (SAR ATR) has been widely deployed in the field of radar signal processing. Nevertheless, these methods are frequently regarded as black-box models due to the limited visual interpretation of their internal feature representation and parameter organization. In this paper, we propose an innovative approach named C-RISE, which builds upon the RISE algorithm to provide a post-hoc interpretation technique for black-box models used in SAR Images Target Recognition. C-RISE generates saliency maps that effectively visualize the significance of each pixel. Our algorithm outperforms RISE by clustering masks that capture similar fusion features into distinct groups, enabling more appropriate weight distribution and increased focus on the target area. Furthermore, we employ Gaussian blur to process the masked area, preserving the original image structure with optimal consistency and integrity. C-RISE has been extensively evaluated through experiments, and the results demonstrate superior performance over other interpretation methods based on perturbation when applied to neural networks for SAR image target recognition. Furthermore, our approach is highly robust and transferable compared to other interpretable algorithms, including white-box methods.

Affective computing in general and human activity and intention analysis in particular, is a rapidly growing field of research. Head pose and emotion changes, present serious challenges when applied to player’s training and ludology experience in serious games or analysis of customer satisfaction regarding broadcast and web services or monitoring a driver’s attention. Given the increasing prominence and utility of depth sensors, it is now feasible to perform large-scale collection of three-dimensional (3D) data for subsequent analysis. Discriminative random regression forests was selected in order to rapidly and accurately estimate head pose changes in unconstrained environment. In order to complete the secondary process of recognising four universal dominant facial expressions (happiness, anger, sadness and surprise), emotion recognition via facial expressions (ERFE) was adopted. After that, a lightweight data exchange format (JavaScript Object Notation-JSON) is employed, in order to manipulate the data extracted from the two aforementioned settings. Motivated by the need of generating comprehensible visual representations from different sets of data, in this paper we introduce a system capable of monitoring human activity through head pose and emotion changes, utilising an affordable 3D sensing technology (Microsoft Kinect sensor).

In order to improve real-time performance of gesture recognition by micro-Doppler map of mmWave Radar, the point cloud based gesture recognition for mmWave Radar is proposed in this paper.Two steps are carried out for mmWave Radar based gesture recognition. The first step is to estimate the point cloud of the gestures by 3D-FFT and the peak grouping. The second step is to train the TRANS-CNN model by combining the multi-head self-attention and the 1D-convolutional network, so as to extract the features in the point cloud data at a deeper level to categorize the gestures. In the experiments, TI mmWave Radar sensor IWR1642 is used as benchmark to evaluate the feasibility of the proposed approach. The results show that the accuracy of the gesture recognition reaches 98.5%. In order to prove the effectiveness of our approach, a simply 2Tx2Rx Radar sensor is developed in our lab and the accuracy of recognition reaches 97.1%. The results show that our proposed gesture recognition approach achieves the best performance in real-time with limited training data, in comparison with the existing methods.

Abstract:The objects of fine-grained image categories(e.g., bird species) are various subclass under different categories. Because the differences between subclass are very subtle and most of them are concentrated in multiple local areas, the task of fine-grained image recognition is very challenging. At the same time, some fine-grained networks tend to focus on a certain region when judging the target category, resulting in the lack of other auxiliary regional features. To this end, Inception V3 is used as the backbone network, and an enhanced and complementary fine-grained image classification network is designed. While adopting the method of reinforcement learning to obtain more detailed fine grain image features, the complementary network can obtain the complementary discriminant area of the target through the method of attention erasure to increase the network's perception of the overall target. Finally, experiments are conducted on CUB-200-2011, FGVC Aircraft and Stanford dogs three open datasets. The experimental results show that the proposed model has better performance.

Recent advancements in Artificial Intelligence (AI), Deep Learning (DL), and computer vision have revolutionized various industrial processes through image classification and object detection. State-of-the-art Optical Character Recognition (OCR) and Object Detection (OD) technologies, such as YOLO and PaddleOCR, have emerged as powerful solutions for addressing challenges in recognizing textual and non-textual information on printed stickers. However, a well-established framework integrating these cutting-edge technologies for industrial applications still needs to be discovered. In this paper, we propose an innovative framework that combines advanced OCR and OD techniques to automate visual inspection processes in an industrial context. Our primary contribution is a comprehensive framework adept at detecting and recognizing textual and non-textual information on printed stickers within a company, harnessing the latest AI tools and technologies for sticker information recognition. Our experiments reveal an overall macro accuracy of 0.88 for the sticker OCR across three distinct patterns. Furthermore, the proposed system goes beyond traditional Printed Character Recognition (PCR) by extracting supplementary information, such as barcodes and QR codes present in the image, significantly streamlining industrial workflows and minimizing manual labor demands.

We propose a novel transfer learning method for speech emotion recognition allowing us to obtain promising results when only few training data is available. With as low as 125 examples per emotion class, we were able to reach a higher accuracy than a strong baseline trained on 8 times more data. Our method leverages knowledge contained in pre-trained speech representations extracted from models trained on a more general self-supervised task which doesn’t require human annotations, such as the wav2vec model. We provide detailed insights on the benefits of our approach by varying the training data size, which can help labeling teams to work more efficiently. We compare performance with other popular methods on the IEMOCAP dataset, a well-benchmarked dataset among the Speech Emotion Recognition (SER) research community. Furthermore, we demonstrate that results can be greatly improved by combining acoustic and linguistic knowledge from transfer learning. We align acoustic pre-trained representations with semantic representations from the BERT model through an attention-based recurrent neural network. Performance improves significantly when combining both modalities and scales with the amount of data. When trained on the full IEMOCAP dataset, we reach a new state-of-the-art of 73.9% unweighted accuracy (UA).

This paper propose an original approach of achieving a Cymatics based visual perception of isolated speech commands. The idea is to smartly combine the effective speech processing and analysis methods with the phenomena of Cymatics. In this context, an effective approach for automatic isolated speech based message recognition is proposed. The incoming speech segment is enhanced by applying the appropriate pre-emphasis filtering, noise thresholding and zero alignment operations. The Mel-Frequency Cepstral coefficients (MFCCs), Delta coefficients and Delta-Delta coefficients are extracted from the enhanced speech segment. Later on, the Dynamic Time Warping (DTW) technique is employed to compare these extracted features with the reference templates. The comparison outcomes are used to make the classification decision. The classification decision is transformed into a methodical excitation. Finally, this excitation is converted into the systematic visual perceptions via the phenomenon of Cymatics. The system functionality is tested with an experimental setup and results are presented. The approach is novel and can be employed in various applications like visual art, encryption, education, archeology, architecture, integration of impaired people, etc.

Speech recognition, also known as automatic speech recognition (ASR), is a technology that enables software to transcribe spoken language into text. However, existing Amharic ASR methods require multiple separate blocks, such as language, acoustic, and pronunciation models with dictionaries, which can be time-consuming and influence performance. This study proposes an approach that replaces much of the speech pipeline with a single recurrent neural network (RNN) architecture. Our proposed architecture is based on a hybrid approach that combines a convolutional neural network (CNN) with a recurrent neural network (RNN) and a connectionist temporal classification (CTC) loss function. We conducted several experiments with noisy audio data that contain 20,000 valid sentences. The model was evaluated using the word error rate (WER) metric, achieving impressive results of 7% WER on noisy data. This approach has significant implications for the field of speech recognition, as it reduces the human effort required to create dictionaries and improves the efficiency and accuracy of ASR systems, making them more practical for real-world applications.

This article provides a general overview of the bio-inspired computing method called DNA-Based Computing (DBC), including its theory and applications. The main theme of DBC is the central dogma of molecular biology (once information of DNA/RNA has got into a protein, it can’t get out again), i.e., DNA to RNA (sequences of four types of nucleotides) and DNA/RNA to protein (sequence of twenty types of amino acids) are allowed, not the reverse ones. Thus, DBC transfers few-element information (DNA/RAN-like) to many-element information (protein-like), solving a given cognitive problem. DBC can take many forms; this article elucidates two main forms, denoted as DBC-1 and DBC-2. Using arbitrary numerical examples, it is shown that DBC-1 can solve the following cognitive problems: “similarity indexing between seemingly different but inherently identical objects” and “recognizing regions of an image separated by a complex boundary.” DBC-2 can solve the following cognitive problem: “pattern recognition when the relevant information is insufficient.” Smart manufacturing-based systems (digital twins and big data analytics) must solve the abovementioned problems to make the manufacturing enablers (machine tools and monitoring systems) more self-reliant and autonomous. Consequently, DBC can improve the cognitive problem-solving ability of smart manufacturing-relevant systems and, thereby, can enhance its biologicalization.

In this paper I tentatively answer 50 questions sampled from a pool of over 10,000 weekly questions formulated by students in a course entitled “The Self”. The questions pertain to various key topics about self-processes, such as self-awareness, self-knowledge, self-regulation, self-talk, self-esteem, and self-regulation. The students’ weekly questions and their answers highlight what is currently know about the self. Answers to the student questions also allow for the identification of some recurrent lessons about the self. Some of these lessons include: all self-processes are interconnected (e.g., prospection depends on autobiography), self-terms must be properly defined (e.g., self-rumination and worry are not the same), inner speech plays an important role in self-processes, controversies are numerous (are animals self-aware?), measurement issues abound (e.g., self-reflection as an operationalization of self-awareness), deficits in some self-processes can have devastating effects (e.g., self-regulatory deficits may lead to financial problems), and there are lots of unknowns about the self (e.g., gender differences in Theory-of-Mind).

Our image recognition system mainly judges whether the left upper limb in the image is the left upper limb or the right upper limb through our deep learning model in the image. The doctor then could give the correct surgical position. From the experimental results, it could be found that the precision rate and recall rate of the intelligent image recognition system proposed in this paper for preventing the upper limb dislocation surgery could reach 98% and 93%, respectively. It proved that our artificial intelligent image recognition system, AIIRS, could indeed assist orthopedic surgeons to prevent the occurrence of left and right dislocation in upper limb surgery. At the same time, this paper also completes the IRB application approval through the prototype experimental results and will conduct the second phase of human trials in the future. It showed that the research results of this paper will be of great benefit and research value to upper limb orthopedic surgery.

In this paper, we continue the research cycle on the properties of convolutional neural network-based image recognition systems and ways to improve noise immunity and robustness [1]. Currently, a popular research area related to artificial neural networks is adversarial attacks. The effect of adversarial attacks on the image is not highly perceptible to the human eye, also it drastically reduces the neural network accuracy. Image perception by a machine is highly dependent on the propagation of high frequency distortions throughout the network. At the same time, a human efficiently ignores high-frequency distortions, perceiving the shape of objects as a whole. The approach proposed in this paper can improve the image recognition accuracy in the presence of high-frequency distortions, in particular, caused by adversarial attacks. The proposed technique makes it possible to measure up the logic of artificial neural network to that of a human, for whom high-frequency distortions are not decisive in object recognition.

Our image recognition system mainly judges whether the left upper limb in the image is the left upper limb or the right upper limb through our deep learning model in the image. The doctor then could give the correct surgical position. From the experimental results, it could be found that the precision rate and recall rate of the intelligent image recognition system proposed in this paper for preventing the upper limb dislocation surgery could reach 98% and 93%, respectively. It proved that our intelligent image recognition system could indeed assist orthopedic surgeons to prevent the occurrence of left and right dislocation in upper limb surgery. At the same time, this paper also completes the IRB application approval through the prototype experimental results and will conduct the second phase of human trials in the future. It showed that the research results of this paper will be of great benefit and research value to upper limb orthopedic surgery.