This paper develops a multi-task learning framework that attempts to incorporate the image structure knowledge to assist image inpainting, which is not well explored in previous works. The primary idea is to train a shared generator to simultaneously complete the corrupted image and corresponding structures --- edge and gradient, thus implicitly encouraging the generator to exploit relevant structure knowledge while inpainting. In the meantime, we also introduce a structure embedding scheme to explicitly embed the learned structure features into the inpainting process, thus to provide possible preconditions for image completion. Specifically, a novel pyramid structure loss is proposed to supervise structure learning and embedding. Moreover, an attention mechanism is developed to further exploit the recurrent structures and patterns in the image to refine the generated structures and contents. Through multi-task learning, structure embedding besides with attention, our framework takes advantage of the structure knowledge and outperforms several state-of-the-art methods on benchmark datasets quantitatively and qualitatively.

In recent years, the emergence of spatiotemporal big data has made the transition of functional identification from the physical dimension to socioeconomic or human activities becoming more common. In the identification of urban functional areas, most studies considered only a single data source and a single division scale, the research results have problems such as low update frequency or incomplete information in a single data set, and overfitting or underfitting in a single spatial resolution. Using taxi trajectory data and point of interest (POI) data as the main data source, this study proposes a multi-scale recursive identification method for urban functional areas based on cross-validation. First, used the dynamic time warping (DTW) algorithm generates a time series similarity matrix, the CA-RFM model combines the clustering algorithm and random forest model is constructed, the model uses a clustering algorithm (K-MEDOIDS) to extract sig-nificant feature regions as input, which are imported into the random forest model for UFZ identification. Then, to overcome the shortcomings of single scale in expressing urban structural characteristics, a recursive model of different levels of urban road networks is established to classify multi-scale functional areas. Finally, cross-validation using the CA-RFM model and POI quantitative identification method, obtains the final identification results of urban functional areas. This paper selects Shenzhen as the study area for the case study, the results show that the com-bination of clustering algorithm and random forest model greatly reduces the error of manual selection of training samples. In addition, the research shows the superiority of the multi-scale recursive identification method that fuses multi-source data and performs cross-validation from two aspects, that is, the division speed of urban functional area identification results is accelerated and the accuracy is improved.

Generative models for images, audio, text and other low-dimension data have achieved great success in recent years. Generating artificial human movements can also be useful for many applications, including improvement of data augmentation methods for human gesture recognition. The object of this research is to develop a generative model for skeletal human movement, allowing to control the action type of generated motion while keeping the authenticity of the result and the natural style variability of gesture execution. We propose to use a conditional Deep Convolutional Generative Adversarial Network (DC-GAN) applied to pseudo-images representing skeletal pose sequences using Tree Structure Skeleton Image format. We evaluate our approach on the 3D-skeleton data provided in the large NTU RGB+D public dataset. Our generative model can output qualitatively correct skeletal human movements for any of its 60 action classes. We also quantitatively evaluate the performance of our model by computing Frechet Inception Distances, which shows strong correlation to human judgement. Up to our knowledge, our work is the first successful class-conditioned generative model for human skeletal motions based on pseudo-image representation of skeletal pose sequences.

Video quality assessment (VQA) plays an important role in video applications for quality evaluation and resource allocation. It aims to evaluate the video quality consistent with the human perception. In this letter, a hierarchical gradient similarity based VQA metric is proposed inspired by the structure of the primate visual cortex, in which visual information is processed through sequential visual areas. These areas are modeled with the corresponding measures to evaluate the overall perceptual quality. Experimental results on the LIVE database show that the proposed VQA metric significantly outperforms the state-of-the-art VQA metrics.

Accurate determination of protein localization, levels, or protein-protein interactions is pivotal for the study of their function, and in situ protein labeling via homologous recombination has emerged as a critical tool in many organisms. While this approach has been refined in various model fungi, the study of protein function in most plant pathogens has predominantly relied on ex-situ or overexpression manipulations. To dissect the molecular mechanisms of development and infection for Verticillium dahliae, a formidable plant pathogen responsible for vascular wilt diseases, we have established a robust, homologous recombination-based in situ protein labeling strategy in this organism. Utilizing Agrobacterium tumefaciens-mediated transformation (ATMT), this methodology facilitates the precise tagging of specific proteins at their C-termini with epitopes, such as GFP and Flag, within the native context of V. dahliae. We demonstrate the efficacy of our approach through the in situ labeling of VdCf2 and VdDMM2, followed by subsequent confirmation via subcellular localization and protein-level analyses. Our findings confirm the applicability of homologous recombination for in situ protein labeling in V. dahliae and suggest its potential utility across a broad spectrum of filamentous fungi. This labeling method stands to significantly advance the field of functional genomics in plant pathogenic fungi, offering a versatile and powerful tool for the elucidation of protein function.

Age-related Macular Degeneration (AMD) is the major cause of elders’ vision loss, early screening and treatment are the most efficient way to reduce the rate of blindness. AI-based methods based on ophthalmic images play great potential for AMD diagnosis. However, low levels of accuracy, robustness, and explainability are challenges for AI approaches applied in clinics. Thus, this study proposed a multi-type of data source fusion method and a multi-model fusion approach for AMD detection. Typical unsupervised (Hierarchical Clustering and K-Means), typical supervised (SVM, VGG-16, and ResNet) methods, and proposed methods (multi-source data fusion-based method and multi-model fusion-based approach) are compared based on Optical Coherence Tomography (OCT), Fundus Autofluorescence (FAF), regular color fundus photography (CFP) and Ultra-Wide field Fundus (UWF) images. Data preprocessing and enhancements of each type of data are discussed. A feature extraction based on unsupervised ML models, feature combination and normalization, and multi-layer perception (MLP) algorithm are involved in the proposed multi-source data fusion-based method. Supervised ML and DL models and a voting mechanism are involved in the multi-model fusion-based approach. Findings show that the proposed methods present a high performance of accuracy and robustness. A real-world UWF database is involved from Shenzhen Aier Hospital. Practical and theoretical contributions are delivered. A reference value for medical diagnosis based on multiple digital images is contributed.

A novel fluorescent probe (C460@Tb-MOFs) was designed and synthesized through encapsulating the fluorescent dye 7-diethylamino-4-methyl coumarin into terbium-based metal-organic framework by a simple ultrasonic impregnation method. It is impressive that this dye-modified metal-organic framework can specifically detect styrene and temperature upon luminescence quenching. The sensing platform of this material exhibit great selectivity, fast response and good cyclability toward styrene detection. It is worth mentioning that the sensing process undergoes a distinct color change from blue to colourless, providing conditions for accurate visual detection of styrene liquid and gas. The significant fluorescence quenching mechanism of styrene toward C460@Tb-MOFs is explored in detail. Moreover, the dye-modified metal-organic framework can also achieve temperature sensing from 298 to 498 K with high relative sensitivity at 498 K. The preparation of functionalized MOFs composites by fluorescent dyes provides an effective strategy for the construction of sensors for multifunctional applications.

In this study, we de novo assembled whole genomes of two small body-sized West Pacific sciaenids (Larimichthys polyactis and Collichthys lucidus) and compared them with published genome data of two closely-related, large body-sized species (Larimichthys crocea and Miichthys miiuy) and one distantly-related, large body-sized outgroup species (Dicentrarchus labrax). The phylogeny constructed using 7,403 single-copy orthologs shared among the five species indicated that L. crocea and L. polyactis diverged about 42 MYA. The two sibling taxa are more closely-related to C. lucidus than M. miiuy. We further identified four growth-related genes (CDHR2, PGC, PTN and PDGFA) that host five diagnostic amino acid variants on body size traits in the fishes, splitting small-body sized L. polyactis and C. lucidus from large-body sized L. crocea, M. miiuy and D. labrax. The results provide new genomic resources and guidelines to facilitate future endeavors in studying functional genomics and developing selective breeding programs for desirable growth traits in sciaenids.

Laminopathies are a spectrum of diseases caused by LMNA mutations. In familial partial lipodystrophy of Dunnigan (FPLD), LMNA plays role in the differentiation and development of adipocytes and lipid metabolism. Changes in LMNA predict not only the differentiation of adipose-derived mesenchymal stem cells (AD-MSCs) but also the transformation of cancer cells. Hence, our in-depth study aimed to identify the molecular connection between disordered lipid metabolism and hepatic carcinogenesis. We first discovered significant positive correlations between pLMNA and two key rate-limiting enzymes in de novo fatty acid synthesis, acetyl-CoA-carboxylase 1 (ACC1) and fatty acid synthase (FASN), in the liver tissue but not in adipose tissue of obese model rats. Moreover, LMNA knockdown (KD) in rat AD-MSCs prevented the differentiation and maturation of adipocytes. To clarify the mechanistic relationship with lipogenesis, gain- and loss-of-function experiments in which functional changes and the related molecular pathways were investigated in a normal hepatocyte line (7701 cells). Adenosine 5'-monophosphate activated protein kinase α (AMPKα) was found to be activated by abnormalities in the LMNA structure under conditions of LMNA deletion, farnesyltransferase inhibitor (FTI) treatment and LMNA mutations associated with clinical FPLD pathogenic phenotype. Active AMPKα could directly phosphorylate ACC1 and thus inhibit lipid synthesis but induced glycolysis in both HCC cells and normal cells. The HCC cells could not survive with LMNA knockout (KO) or even KD. Lonafarnib (an FTI) combined with low-glucose conditions significantly decreased the proliferation of HepG2 and MHCC cells by inhibiting glycolysis and the maturation of prelamin A.

Objective: To investigate the effects and mechanisms of different concentrations of CCCP on mitophagy in human peripheral blood regulatory T cells. Methods: Tregs were isolated, identified and then grouped, treating with CCCP at a concentration of 2.5 μM, 5 μM, 10 μM, 20 μM and 40 μM for 24h in an incubator. Flow cytometry detected the reactive oxygen species (ROS), mitochondrial membrane potential (MMP), mitochondrial quality, and fluorescence microscopy observed the co-localization of mitochondria and lysosomes in each group. Results: The purity of CD4⁺CD25⁺Tregs was (93.36 ± 1.87) %. With the increase of CCCP concentration, the level of ROS gradually increased, while the MMP decreased gradually. About the mitochondria and lysosome fusion, the fluorescence intensity of orange (yellow) was the highest when the concentration of CCCP was in the range of 5-10 μM while decreased with the CCCP concentration continually increasing. The mitochondrial quality decreased with the increase of CCCP concentration. However, there was no significant difference between groups C, D and E. The mitochondrial quality of groups F and G were significantly lower than that of group E. Conclusions: With the concentration of CCCP gradually increased, the level of ROS in Treg cells increased, and MMP decreased, which promoted the mitophagy, mitochondrial quality maintains homeostasis; When ROS accumulated, and MMP decreased significantly, the mitophagy was inhibited, and the mitochondrial quality was significantly decreased.

Abstract: Bacillus subtilis fmb60, which has broad-spectrum antimicrobial activities was isolated from plant straw compost. A hybrid NRPS/PKS cluster was screened from the genome. Sixteen secondary metabolites produced by the gene cluster were isolated and identified using LC-HRMS and NMR. Three lipoamides D–F (1-3) and two amicoumacin derivatives, amicoumacins D, E (4, 5), were identified, and are reported here for the first time. Lipoamides D–F exhibited strong antibacterial activities against harmful foodborne bacteria, with the MIC ranging from 6.25 to 25 µg/mL. Amicoumacin E scavenged 38.8% of ABTS+ radicals at 1 mg/mL. Direct cloning and heterologous expression of the NRPS/PKS and ace gene cluster identified its importance for the biosynthesis of amicoumacins. This study demonstrated that there is a high potential for biocontrol utilization of B. subtilis fmb60, and genome mining for clusters of secondary metabolites of B. subtilis fmb60 has revealed a greater biosynthetic potential for the production of novel natural products than previously anticipated.

Hypoglossal nerve stimulator (HGNS) is a minimally invasive device used for treating obstructive sleep apnea (OSA). The conventional implantable HGNS is a device that consists of a stimuli generator, a breathing sensor, and electrodes connected to the hypoglossal nerve via leads. However, this implant is bulky and causes significant trauma. In this paper, we propose a minimally invasive HGNS based on an electrocardiogram (ECG) sensor and a wireless power transfer (WPT), consisting of a wearable breathing monitor and an implantable stimulator. The breathing external monitor utilizes an ECG sensor to identify abnormal breathing patterns associated with OSA with 88.68$\%$ accuracy, achieved through the utilization of a Convolutional Neural Network (CNN) algorithm. With a skin thickness of 5mm and a receiving coil diameter of 9mm, the power conversion efficiency was measured at 31.8$\%$. The implantable device, on the other hand, is composed of a front-end CMOS Power Management Module (PMM), a Binary Phase Shift Keying (BPSK)-based data demodulator, and a bipolar biphasic current stimuli generator. The PMM, with a silicon area of 0.06 $mm^2$ (excluding pads), demonstrates a power conversion efficiency of 77.5$\%$ when operating at a receiving frequency of 2 MHz. Furthermore, it offers three-voltage options (1.2V, 1.8V, and 3.1V). Within the data receiver component, a low-power BPSK demodulator has been ingeniously incorporated, consuming only 42 $\mu$W when supplied with a voltage of 0.7V. The performance is achieved through the implementation of the self-biased phase-locked loop (PLL) technique. The stimuli generator delivers biphasic constant currents, providing a 5-bit programmable range spanning from 0 to 2.4 mA. The functionality of proposed ECG and WPT-based HGNS was validated, representing a highly promising solution for the effective management of OSA, all while minimizing trauma and space requirements.

Macula fovea detection is a crucial prerequisite towards screening and diagnosing macular diseases. Without early detection and proper treatment, any abnormality involving the macula may lead to blindness. However, with the ophthalmologist shortage and time-consuming artificial evaluation, neither accuracy nor effectiveness of the diagnose process could be guaranteed. In this project, we proposed a deep learning approach on ultra-widefield fundus (UWF) images for macula fovea detection. This study collected 2300 ultra-widefield fundus images from Shenzhen Aier Eye Hospital in China. Methods based on U-shape network (Unet) and Fully Convolutional Networks (FCN) are implemented on 1800 (before amplifying process) training fundus images, 400 (before amplifying process) validation images and 100 test images. Three professional ophthalmologists were invited to mark the fovea. A method from the anatomy perspective is investigated. This approach is derived from the spatial relationship between macula fovea and optic disc center in UWF. A set of parameters of this method is set based on the experience of ophthalmologists and verified to be effective. Results are measured by calculating the Euclidean distance between proposed approaches and the accurate grounded standard, which is detected by Ultra-widefield swept-source optical coherence tomograph (UWF-OCT) approach. Through a comparation of proposed methods, we conclude that, deep learning approach of Unet outperformed other methods on macula fovea detection tasks, by which outcomes obtained are comparable to grounded standard method.

In this research, catalytic steam reforming acetic acid derived from the aqueous portion of bio-oil for hydrogen production was investigated by using different Ni/ATC (Attapulgite Clay) catalysts prepared by precipitation, impregnation and mechanical blending methods. The fresh and reduced catalysts were characterized by XRD, N₂ adsorption-desorption, TEM and H₂-TPR. The comprehensive results demonstrated that the interaction between active metallic Ni and ATC carrier was significantly improved in Ni/ATC catalyst prepared by precipitation method, and in which the mean Ni particle size was the smallest (~13 nm) resulted in the highest metal dispersion (7.5%). The catalytic performance of the three catalysts was evaluated through the process of steam reforming of acetic acid in a fixed-bed reactor under atmospheric pressure at two different temperatures, such as 550 ℃ and 650 ℃. Results showed that the Ni/ATC (PM-N/ATC) prepared by precipitation method, achieved the highest H₂ yield of ~82% and little lower acetic acid conversion efficiency of ~85% than that (~95%) of Ni/ATC (IM-NATC) prepared by impregnation method. In addition, the deactivation catalysts after reaction for 4 h were analyzed by XRD, TGA-DTG and TEM, which demonstrated that the catalyst deactivation was not caused by the amount of carbon deposition, but owed to the significant agglomeration and sintering of Ni particles in the carrier.