Explore the vast range of titles available.

Drug discovery remains a costly and time‐intensive process, where accurate identification of drug associations is critical for therapeutic development. Existing computational approaches predominantly rely on sequence‐derived or 2D molecular representations, often overlooking the intrinsic 3D complexity of small molecules. Here, a deep learning framework is presented that directly learns from 3D molecular spatial visual information, capturing geometric, topological, and stereochemical features from spatial renderings. By integrating this spatial information with traditional molecular descriptors, unified multi‐perspective representations are constructed that better reflect molecular structure and function. Across benchmark tasks involving drug–microRNA, drug–drug, and drug–protein interaction prediction, this model consistently outperforms conventional fingerprint‐based baselines. Interpretability analyses show that the model attends to biologically relevant substructures, highlighting the value of 3D molecular spatial visual information in molecular recognition. These findings demonstrate the potential of spatially informed learning to enhance predictive performance and provide mechanistic insights in computational drug discovery. A deep learning framework called MolVisGNN is proposed to fuse 3D molecular visual information of drugs with multi‐source features, which proves the importance of 3D molecular visual information of drugs and the advancedness of this model in the field of drug discovery, and provides a reference for how to more comprehensively express small molecule drugs in deep learning in the future.

Organizations maintain process models that describe or prescribe how cases (e.g., orders) are handled. However, reality may not agree with what is modeled. Conformance checking techniques reveal and diagnose differences between the behavior that is modeled and what is observed. Existing conformance checking approaches tend to focus on the control-flow in a process, while abstracting from data dependencies, resource assignments, and time constraints. Even in those situations when other perspectives are considered, the control-flow is aligned first, i.e., priority is given to this perspective. Data dependencies, resource assignments, and time constraints are only considered as “second-class citizens”, which may lead to misleading conformance diagnostics. For example, a data attribute may provide strong evidence that the wrong activity was executed. Existing techniques will still diagnose the data-flow as deviating, whereas our approach will indeed point out that the control-flow is deviating. In this paper, a novel algorithm is proposed that balances the deviations with respect to all these perspectives based on a customizable cost function. Evaluations using both synthetic and real data sets show that a multi-perspective approach is indeed feasible and may help to circumvent misleading results as generated by classical single-perspective or staged approaches.

Around the globe, all developing and developed economies focus on renewable energy sources for clean and sustainable economic growth and aim to decrease carbon emission to improve the climatic condition. This study aims to evaluate the feasible renewable energy source in Pakistan and study the criteria based on the investigation of wind energy, biomass energy, geothermal and solar energy, and their effects on Pakistan’s climate. As renewable energy, clean and sustainable production is only produced from renewable energy sources; the current paper bears the viability of five primary renewable energy sources in Pakistan. This article established an advanced fuzzy (AHP) with a multi-perspective approach based on the expert assessment to select the efficient renewable energy in Pakistan. Further in this study, we utilized the DEA model to investigate the relative efficacy of renewable energy. This study indicated that wind energy is the efficient renewable energy source in Pakistan, which showed an efficiency score of 0. 1000. Our results further suggested that wind energy’s next option for renewable energy solar best for renewable production in Pakistan. The results further exposed that the solar is the second-highest energy source with an efficiency score of 0. 777. Moreover, geothermal gets a less efficient score, that of 0.198. Further, our study showed that the economic benefits (EBs) and environmental implication criteria are significant in this research. This article gives a strong recommendation for a future wind energy source in Pakistan. Our results suggest that the best renewable energy source in the future is wind energy for the sustainable and economic development of Pakistan and other developing countries.

Semi-supervised learning (SSL) is an effective method for medical image segmentation as it alleviates the dependence on clinical pixel-level annotations. Among the SSL methods, pseudo-labels and consistency regularization play a key role as the dominant paradigm. However, current consistency regularization methods based on shared encoder structures are prone to trap the model in cognitive bias, which impairs the segmentation performance. Furthermore, traditional fixed-threshold-based pseudo-label selection methods lack the utilization of low-confidence pixels, making the model’s initial segmentation capability insufficient, especially for confusing regions. To this end, we propose a multi-perspective dynamic consistency (MPDC) framework to mitigate model cognitive bias and to fully utilize the low-confidence pixels. Specially, we propose a novel multi-perspective collaborative learning strategy that encourages the sub-branch networks to learn discriminative features from multiple perspectives, thus avoiding the problem of model cognitive bias and enhancing boundary perception. In addition, we further employ a dynamic decoupling consistency scheme to fully utilize low-confidence pixels. By dynamically adjusting the threshold, more pseudo-labels are involved in the early stages of training. Extensive experiments on several challenging medical image segmentation datasets show that our method achieves state-of-the-art performance, especially on boundaries, with significant improvements.

In order to study the trend of criminal behavior from the perspective of temporal and spatial distribution, this paper adopts the time series decomposition method and the nearest neighbor index method to obtain the temporal and spatial distribution patterns of cases. A multi-view simultaneous convolution algorithm is designed to extract and fuse features from three perspectives: time, space and type to predict crime trends. Based on the City B law enforcement database, a study was conducted to explore the temporal heterogeneity of burglary crimes by combining two-dimensional color matrix plots and multi-timescale trend plots. Standardized crime rates and spatial autocorrelation tests are used to reveal the spatial aggregation patterns of burglary crimes. Spatio-temporal features are fused based on type nodes to dissect the potential patterns of burglary crimes. The prediction results are evaluated in terms of performance and prevention suggestions are made for burglary crime trends. The results indicate that there are roughly three levels of time periods in the distribution of crimes in a day. The lowest number of burglary incidents is from 0-6 hours, the relatively high number is from 6-18 hours, and the highest number of hours is at 18-24 hours. In view of the spatial and temporal characteristics of burglary, the strategy of “multi-point joint defense and cooperative action” can effectively prevent and reduce the occurrence of burglary cases by enhancing patrols and strengthening the security of key areas.

China's urban population will increase by 268 million from 2010 to 2030, with the consumption of a large number of resource‐intensive products. Quantitative analysis of the environmental impacts (water, energy and carbon) of urban agglomerations can make trade‐offs among water conservation, energy use, climate change mitigation, and urban development. In this study, a multi‐layer water‐energy‐carbon production path analysis (MWPPA) model is developed for identifying the key final demands, sectors and supply chain paths of the Pearl River Delta urban agglomeration (PUA). Results show that, water, energy and carbon‐emission intensities respectively reduced by 27.3%, 35.6% and 27.6% in 2015, compared to the levels in 2012. More than half of the water‐energy‐carbon (WEC) footprints are export‐driven, where Guangzhou, Shenzhen and Foshan dominate the WEC footprints of PUA. Results also disclose that Shenzhen is the main recipient of water‐energy, while Jiangmen and Huizhou are the main providers of water and energy, respectively. Policy makers are suggested that each industry actively integrate into global value chains in order to leverage its comparative advantage, and Huizhou should take full advantage of its fossil base to form a complete industry chain from the R&D end to the production end around the energy industry. Plain Language Summary Not only do urban areas consume large amounts of water and energy, they are also the specific implementation units of carbon reduction policies. The United Nations Sustainable Development goals (UN SDGs) make it clear that water conservation, energy access, climate change mitigation, and urbanization development are important parts of its agenda. As one of the most developed urban agglomerations in China, the Pearl River Delta urban agglomeration (PUA) is also the main consumer of water, energy and carbon (WEC). This study reveals that more than half of the WEC footprints are export‐driven, and Guangzhou, Shenzhen and other developed economies dominate the WEC footprints of PUA. Compared to 2012, the consumption intensity of WEC was reduced in 2015. The results also find that light industry and equipment manufacture play key roles in the WEC system. Results suggest that greener production needs to be adopted not only within but also outside of urban agglomerations, while individual cities need to actively promote the integration of each industry into global supply chains. Key Points A multi‐layer water‐energy‐carbon production path analysis model is developed Exports drive more than half of water, energy and carbon footprints in the Pearl River Delta In the Pearl River Delta, cities of Jiangmen and Huizhou supply the most water and energy, and Huizhou and Dongguan provide the most carbon

With the large number of high-resolution images now being acquired, high spatial resolution (HSR) remote sensing imagery scene classification has drawn great attention but is still a challenging task due to the complex arrangements of the ground objects in HSR imagery, which leads to the semantic gap between low-level features and high-level semantic concepts. As a feature representation method for automatically learning essential features from image data, convolutional neural networks (CNNs) have been introduced for HSR remote sensing image scene classification due to their excellent performance in natural image classification. However, some scene classes of remote sensing images are object-centered, i.e., the scene class of an image is decided by the objects it contains. Although previous methods based on CNNs have achieved comparatively high classification accuracies compared with the traditional methods with handcrafted features, they do not consider the scale variation of the objects in the scenes. This makes it difficult to directly utilize CNNs on those remote sensing images belonging to object-centered classes to extract features that are robust to scale variation, leading to wrongly classified scene images. To solve this problem, scene classification based on a deep random-scale stretched convolutional neural network (SRSCNN) for HSR remote sensing imagery is proposed in this paper. In the proposed method, patches with a random scale are cropped from the image and stretched to the specified scale as the input to train the CNN. This forces the CNN to extract features that are robust to the scale variation. Furthermore, to further improve the performance of the CNN, a robust scene classification strategy is adopted, i.e., multi-perspective fusion. The experimental results obtained using three datasets—the UC Merced dataset, the Google dataset of SIRI-WHU, and the Wuhan IKONOS dataset—confirm that the proposed method performs better than the traditional scene classification methods.

Anomaly detection is a critical problem in the manufacturing industry. In many applications, images of objects to be analyzed are captured from multiple perspectives which can be exploited to improve the robustness of anomaly detection. In this work, we build upon the deep support vector data description algorithm and address multi-perspective anomaly detection using three different fusion techniques, i.e., early fusion, late fusion, and late fusion with multiple decoders. We employ different augmentation techniques with a denoising process to deal with scarce one-class data, which further improves the performance (ROC AUC =80%). Furthermore, we introduce the dices dataset, which consists of over 2000 grayscale images of falling dices from multiple perspectives, with 5% of the images containing rare anomalies (e.g., drill holes, sawing, or scratches). We evaluate our approach on the new dices dataset using images from two different perspectives and also benchmark on the standard MNIST dataset. Extensive experiments demonstrate that our proposed multi-perspective approach exceeds the state-of-the-art single-perspective anomaly detection on both the MNIST and dices datasets. To the best of our knowledge, this is the first work that focuses on addressing multi-perspective anomaly detection in images by jointly using different perspectives together with one single objective function for anomaly detection.

The objective of 3D hand pose estimation (HPE) based on depth images is to accurately locate and predict keypoints of the hand. However, this task remains challenging because of the variations in hand appearance from different viewpoints and severe occlusions. To effectively address these challenges, this study introduces a novel approach, called the multi-perspective cue-aware joint relationship representation for 3D HPE via the Swin Transformer (MPCTrans, for short). This approach is designed to learn multi-perspective cues and essential information from hand depth images. To achieve this goal, three novel modules are proposed to utilize features from multiple virtual views of the hand, namely, the adaptive virtual multi-viewpoint (AVM), hierarchy feature estimation (HFE), and virtual viewpoint evaluation (VVE) modules. The AVM module adaptively adjusts the angles of the virtual viewpoint and learns the ideal virtual viewpoint to generate informative multiple virtual views. The HFE module estimates hand keypoints through hierarchical feature extraction. The VVE module evaluates virtual viewpoints by using chained high-level functions from the HFE module. Transformer is used as a backbone to extract the long-range semantic joint relationships in hand depth images. Extensive experiments demonstrate that the MPCTrans model achieves state-of-the-art performance on four challenging benchmark datasets.

Distracted driving continues to be a major contributor to road accidents, highlighting the growing research interest in advanced driver monitoring systems for enhanced safety. This paper seeks to improve the overall performance and effectiveness of such systems by highlighting the importance of recognizing the driver’s activity. This paper introduces a novel methodology for assessing driver attention by using multi-perspective information using videos that capture the full driver body, hands, and face and focusing on three driver tasks: distracted actions, gaze direction, and hands-on-wheel monitoring. The experimental evaluation was conducted in two phases: first, assessing driver distracted activities, gaze direction, and hands-on-wheel using a CNN-based model and videos from three cameras that were placed inside the vehicle, and second, evaluating the multi-task fusion algorithm, considering the aggregated danger score, which was introduced in this paper, as a representation of the driver’s attentiveness based on the multi-task data fusion algorithm. The proposed methodology was built and evaluated using a DMD dataset; additionally, model robustness was tested on the AUC_V2 and SAMDD driver distraction datasets. The proposed algorithm effectively combines multi-task information from different perspectives and evaluates the attention level of the driver.