The success of deep learning (DL) in offline brain-computer interfaces (BCIs) has not yet translated into efficient online applications. This is due to three primary challenges. First, most DL solutions are designed for offline decoding, making the transition to online decoding complex and unclear. Second, the use of sliding windows in online decoding increases the computational complexity of DL training. Third, DL models typically require large amounts of training data, which are often scarce in BCI applications. To address these issues and enable real-time decoding, even across different subjects without calibration data, we first introduce a novel approach called real-time adaptive pooling (RAP). RAP modifies the pooling layers of existing offline DL models towards the online decoding requirements. Additionally, it significantly reduces the computational demand during training by jointly decoding consecutive windows. To reduce the amount of training data required, our approach leverages different levels of domain adaptation. We show how different settings enable different adaptation solutions. The results demonstrate that our approach is both powerful and can be calibration-free, providing a robust and practical solution for real-time BCI applications. These findings pave the way for the development of co-adaptive and highly efficient DL-based online BCI systems.

Infrared and visible image fusion technologies are used to characterize the same scene by diverse modalities. However, most existing deep learning-based fusion methods are designed as symmetric networks, which ignore the differences between modal images and lead to the source image information loss during feature extraction. In this paper, we propose a new fusion framework for the different characteristics of infrared and visible images. Specifically, we design a dual-stream asymmetric network with two different feature extraction networks to extract infrared and visible feature maps respectively. The transformer architecture is introduced in the infrared feature extraction branch, which can force the network to focus on the local features of infrared images while still obtaining their contextual information. And the visible feature extraction branch uses residual dense blocks to fully extract the rich background and texture detail information of visible images. In this way, it can provide better infrared targets and visible details for the fused image. Experimental results on multiple datasets indicate that DSA-Net outperforms state-of-the-art methods in both qualitative and quantitative evaluations. In addition, we also apply the fusion results to the target detection task, which indirectly demonstrates the fusion performances of our method.

The hybrid personal cooling system (HPCS) consisted of ventilation fans and phase change materials (PCMs) covered with insulation pads is a promising wearable cooling system to mitigate heat strain and heat-related illnesses of occupational workers with heavy labor in hot environments. Effects of clothing characteristics (e.g., thermal resistance of insulation pads, latent heat and melting temperature of PCMs) on the thermal performance of the HPCS have been investigated in detail in our previous study. Apart from the aforementioned factors, environmental conditions, i.e., environmental temperature and relative humidity, also significantly affect the thermal performance of the HPCS. In this paper, a numerical parametric study was performed to investigate the effects of the environmental temperature and relative humidity (RH) on the thermal management of the HPCS. Five levels of air temperature under environmental RH=50% were chosen (i.e., 32, 34, 36, 38 and 40 ºC) to study the impact of environmental temperature on the HPCS’s cooling performance. In addition, four levels of environmental RH at ambient temperatures of 36 and 40 ºC were selected (i.e., 30, 50, 70 and 90%) to examine the effect of RH on cooling performance of the HPCS. Results show that high environmental temperatures could accelerate the PCM melting process and thereby weaken the cooling performance of HPCS. In the moderately hot environment (36 °C), the HPCS presented good cooling performance with the maximum core temperature at around 37.5 °C during excise when the ambient RH≤70%, whereas good cooling performance could be only seen under RH≤50% in the extremely hot environment (40 °C). Thus, it may be concluded that the maximum environmental RH for the HPCS exhibiting good cooling performance decreases with the increase in the environmental temperature.

Marine soft corals are prolific sources of various natural products that have served as a wealthy reservoir of diverse chemical scaffolds for new drug leads. The genus Litophyton contains almost 100 species but only a small part of them had been classified and chemically investigated, which calls for more attentions from the global researchers. In the current work, 109 secondary metabolites have been discussed based on published data from 1975 to July, 2023, covering a period of near five decades. The studied species of the genus Litophyton inhabited various tropical and temperate regions and afforded a variety of biologically active natural products such as terpenes, steroids, alkaloids, and lipids. A wide spectrum of pharmacological effects of these compounds had been evaluated, including cytotoxic, anti-HIV, antibacterial, antifungal, anti-malarial, antifeedant, molluscicidal, PTP1B inhibitory, and insect growth inhibitory activities. This review aims to provide an up-to-date literature survey and comprehensive insight into chemical structures, taxonomical distributions, and biological activities of the reported metabolites from the title genus whenever available.

Construction safety is especially important because the construction industry is so important to a country's development. Significant research and practice have been conducted to mitigate potential risks during construction and improve worker efficiency. With the rapid advancement of cognitive neuroscience and the incorporation of medical technology in recent years, various wearable monitoring devices have been widely used in the construction field for real-time monitoring of workers' physical and mental status. Among these, the use of EEG (Electroencephalogram) in construction environment research allows researchers to gain insight into the physical and mental states of construction workers while performing construction tasks. This review introduces EEG technology and portable EEG devices, followed by an overview of their use in both monitoring workers' adverse reactions and identifying hazards on construction sites, providing an effective guide for EEG research in the construction field and on-site safety management.

Nanostructured transition metal nitrides (TMNs) have been considered a promising substitute for precious metal catalysts toward ORR due to their multi-electron orbitals, metallic properties, and low cost. To design TMNs catalysts with high catalytic activity toward ORR, the intrinsic features of the influencing factor on the catalytic activity toward ORR of nanostructured TMNs need to be investigated. In the paper, titanium nitride (TiN), zirconium nitride (ZrN), and hafnium nitride (HfN) nanoparticles (NPs) are high-efficiency and one-step synthesized by the direct current arc plasma. TiN, ZrN, and HfN NPs with an oxidation layer are applied as the catalysts of hybrid sodium-air batteries (HSABs). The effect of composition and structure attributes of TMNs on ORR catalysis is defined as follows. (ⅰ) Composition effect. With the increase of the oxygen content, the catalytic ORR capability of TMNs decreases progressively due to the reduction of oxygen adsorption capacity; (ⅱ) Structure effect. The redistribution of the density of states (DOS) of ZrN indicates higher ORR activity than TiN and HfN. HSABs with ZrN exhibit an excellent cyclic stability up to 137 cycles (about 140 h), an outstanding rate performance, and a specific capacity of 2817 mAh·g−1 at 1.0 mA·cm−2.

Based on the marine natural products piperafizine B, X334 and our previously reported compound 4m, fourteen novel 3,6-diunsaturated 2,5-diketopiperazine (2,5-DPK) derivatives (1-2, 4-6, 8-16), together with two known ones (3 and 7), were designed and synthesized as anticancer agents against cell lines A549 and Hela. The MTT assay results showed that the derivatives 6, 8-12 and 15 had moderate to good anticancer capacities with the IC50 values ranging from 0.7 to 8.9 μM. Among them, compound 11 with naphthalen-1-ylmethylene and 2-methoxybenzylidene functions at the 3- and 6-positions of 2,5-DPK ring, respectively, displayed good inhibitory activities to both cancer cells A549 (IC50 = 1.2 μM) and Hela (IC50 = 0.7 μM). It could also induce apoptosis and obviously block the cell cycle progression at G2/M phase in both cells at 1.0 μM. The electron withdrawing functions might not be favorable for the derivatives with high anticancer activities. Additionally, compared to piperafizine B and X334, these semi-N-alkylated derivatives have high liposolubilities (> 1.0 mg•mL-1). Compound 11 can be further developed aiming to the discovery of novel anticancer candidate.

Incorporating knowledge graphs as auxiliary information to enhance recommendation systems can improve the representations learning of users and items. Recommendation methods based on knowledge graph can introduce user-item interaction learning into the item graph, focusing only on learning the node vector representations within a single graph; alternatively, they can treat user-item interactions and item graphs as two separate graphs and learn from each graph individually. Learning from two graphs has natural advantages in exploring original information and interaction information, but faces two main challenges: 1) in complex graph connection scenarios, how to adequately mine the self-information of each graph, and 2) how to merge interaction information from the two graphs while ensuring that user-item interaction information predominates. Existing methods do not thoroughly explore the simultaneous mining of self-information from both graphs and effective interaction information, leading to the loss of valuable insights. Considering the success of contrastive learning in mining self-information and auxiliary information, this paper proposes a dual-graph contrastive learning recommendation method based on knowledge graphs (KGDC) to explore a more accurate representations of users and items in recommendation systems based on external knowledge graphs. In the learning process within the self-graph, KGDC has strengthened and represented the information of different connecting edges in both graphs, and extracted the existing information more fully. In interactive information learning, KGDC reinforces the interaction relationship between users and items in the external knowledge graph, realizing the leading role of the main task. We have conducted a series of experiments on three standard datasets, and the results show that the proposed method can achieve better results.

Accelerating, exact, explicit and simple solutions of relativistic hydrodynamics allow for a simple and natural description of highly relativistic p+p collisions. These solutions yield a finite rapidity distribution, thus they lead to an advanced estimate of the initial energy density of high energy collisions. We show that such an advanced estimate yields an initial energy density in $\sqrt{s}=7$ and 8 TeV p+p collisions at LHC around or above the critical energy density from lattice QCD, and a corresponding initial temperature above the critical temperature from QCD and the Hagedorn temperature. This suggests that the collision energy of the LHC corresponds to a large enough initial energy density to create a non-hadronic perfect fluid even in pp collisions. %We also show, that several times the %critical energy density may have been reached in high multiplicity events, hinting at a non-hadronic medium created in %high multiplicity $\sqrt{s}=7$ and 8 TeV p+p collisions.

To improve the current density distribution and electromagnetic performance of air-core inductors, a structural optimization method combining BP neural network and genetic algorithm was proposed for the study of axial and radial spiral multi-winding inductors. The Monte Carlo method was used to extract the structural size samples of the inductors, and the training dataset was obtained through finite element calculation of electromagnetic fields. Based on BP neural networks, nonlinear mapping models between the inductance value, volumetric inductance density, current distribution non-uniformity coefficient, and inductor structural parameters were constructed. The sensitivity analysis of the inductor inductance value affected by structural parameters was conducted through the Sobol index calculation. Using the current distribution non-uniformity coefficient as the fitness function and the volumetric inductance density as the constraint condition, the genetic algorithm was applied to optimize the structural parameters of the inductor globally. The optimization results were verified through finite element comparison. The results show that, under the requirement of satisfying the volumetric inductance density, the current distribution non-uniformity coefficient of AHI type inductor was reduced by 4.57% compared with the best sample in the sampling, while that of RHI type inductor was reduced by 5.33%, demonstrating the practicality of the BP-GA joint algorithm in the structural optimization design of inductors.

As a crucial part of head protection equipment, research on the material and structural application of helmet liners has always been a hot topic in the field of helmet manufacturing. This paper aims to introduce the most common helmet liner materials, including traditional polystyrene, polyethylene, polypropylene, and others, as well as emerging materials such as anisotropic materials, and polymer nanocomposites. In addition, the design principles of helmet liner structure are discussed, including the evolution of multi-layer structure, integration of geometric irregular bubbles, etc., to enhance energy absorption capacity. Advanced manufacturing processes such as additive manufacturing technology are introduced to fabricate complex structures. Finally, the application of biomimetic structures in helmet liner design is explored. For example, the structure of biological tissue can be used to design a helmet liner structure with superior energy absorption performance. On this basis, the idea of bionic structures is extended to the combination of plant stalks and animal skeletal structures, and additive manufacturing technology is used to significantly reduce energy loss during elastic yield energy absorption, thus designing a reusable helmet. This provides a promising research direction for future helmet liner materials and structural applications.