Traditional nursing activity recognition relies on manual observation and documentation, which are timeconsuming and error-prone. This paper presents CareAttenNet, a deep learning framework integrating adaptive feature selection, correlation-aware processing, and temporal attention mechanisms for automated nursing activity recognition from wearable sensor data. We evaluated the framework using the SONAR dataset comprising 70-dimensional sensor features from 14 healthcare professionals performing 23 nursing activities, totaling$7,631,843$temporal measurements. CareAttenNet achieved 77.36 % validation accuracy and 60.00 % test accuracy, outperforming baseline architectures including CNN-LSTM (57.92 %), Correlation-Aware CNN$(54.62 \%)$, and Feature-Selective Network$(52.86 \%)$. Ablation studies revealed temporal attention as the most effective component (78.33 % test accuracy), while feature selection combined with temporal attention achieved 77.40 % accuracy. However, combining all components resulted in performance degradation, indicating complex negative interactions between architectural modules. These findings provide insights into multi-modal sensor fusion challenges and establish a foundation for intelligent healthcare monitoring systems.

Psoriatic arthritis (PsA) is a complex inflamma-tory disease and a more severe condition of psoriasis (PsO), frequently accompanied by skin and nail lesions. Early differentiation between these conditions is crucial for appropriate treatment selection and improved patient outcomes. Therefore, this study assesses the contribution of the unaffected fingernail indicators and presents PsACasNet, a multi-modal cascaded fusion neural network for effective classification of PsA out of PsO. The framework integrates the clinical phenotypes, protein structural features, and morphology of the unaffected fingernails of patients. PsACasNet implements a biology-driven cascade fusion pathway from molecular (spectral) to morpho-logical (SEM image), and then to clinical indicators. Through comprehensive ablation studies evaluating seven different modal combinations, we demonstrate the effectiveness of our model, which follows a biology-driven molecular-to-morphological-to-clinical fusion pathway. Experimental results show that dual-modal combinations of Clinical+Spectral and Spectral+Image achieve optimal performance (92.0% accuracy), while spectral features alone reach 84.0% accuracy, highlighting the discrimi-native power of molecular characteristics. The comparison with other traditional machine-learning methods demonstrates the efficiency of our multi-modal fusion strategies. Moreover, through the interpretability analysis of the model, we found that the morphological and protein characteristics of the uninvolved nails have made outstanding contributions to the diagnosis of PsA and can serve as effective biomarkers to promote diagnosis.

Abstract Self-supervised facial representation learning (SFRL) methods, especially contrastive learning (CL) methods, have been increasingly popular due to their ability to perform face understanding without heavily relying on large-scale well-annotated datasets. However, analytically, current CL-based SFRL methods still perform unsatisfactorily in learning facial representations due to their tendency to learn pose-insensitive features, resulting in the loss of some useful pose details. This could be due to the inappropriate positive/negative pair selection within CL. To conquer this challenge, we propose a Pose-disentangled Contrastive Facial Representation Learning (PCFRL) framework to enhance pose awareness for SFRL. We achieve this by explicitly disentangling the pose-aware features from non-pose face-aware features and introducing appropriate sample calibration schemes for better CL with the disentangled features. In PCFRL, we first devise a pose-disentangled decoder with a delicately designed orthogonalizing regulation to perform the disentanglement; therefore, the learning on the pose-aware and non-pose face-aware features would not affect each other. Then, we introduce a false-negative pair calibration module to overcome the issue that the two types of disentangled features may not share the same negative pairs for CL. Our calibration employs a novel neighborhood-cohesive pair alignment method to identify pose and face false-negative pairs, respectively, and further help calibrate them to appropriate positive pairs. Lastly, we devise two calibrated CL losses, namely calibrated pose-aware and face-aware CL losses, for adaptively learning the calibrated pairs more effectively, ultimately enhancing the learning with the disentangled features and providing robust facial representations for various downstream tasks. In the experiments, we perform linear evaluations on four challenging downstream facial tasks with SFRL using our method, including facial expression recognition, face recognition, facial action unit detection, and head pose estimation. Experimental results show that PCFRL outperforms existing state-of-the-art methods by a substantial margin, demonstrating the importance of improving pose awareness for SFRL. Our evaluation code and model will be available at https://github.com/fulaoze/CV/tree/main .

Video-based facial expression recognition (VFER) is challenging due to variations caused by cultural background and expression camouflage. To tackle these problems, researchers introduced eye movement signals to complement visual information. However, existing methods either require expensive devices to capture high-quality eye movements or can only extract low-quality eye movements visually, making them ineffective in the real world. To address this, we propose an eye movement-instructed VFER (EM-VFER) that leverages high-quality eye movements to instruct the visual learning, obtaining robust performance without requiring costly devices during inference. Specifically, our EM-VFER operates in two stages: the high-quality eye movement pre-training stage and the eye movement-instructed video fine-tuning stage. In the pre-training, we compile an Eye-behavior-aided Multimodal Emotion Recognition (EMER) dataset and use it to train a multimodal Transformer. During the fine-tuning, we propose a novel progressive eye movement-instructed learning to take better advantage of the prior knowledge about high-quality eye movement signals from EMER. The instructed fine-tuning model could then make more robust predictions on downstream facial expression datasets. We evaluate our approach on three macroexpression datasets (DFEW, MAFW and Aff-wild2) and two micro-expression datasets (CASME III and CASME II). The results demonstrate that EM-VFER significantly outperforms existing methods. The code will be available.

Demosaicking is a critical process in the digital imaging pipeline, tasked with reconstructing full-color images from sampled data captured by R/G/B color sensors. The challenge arises from two-thirds of the pixel data being missing, which complicates the task of accurate reconstruction. Recent deep learning-based solutions have yielded considerable advancements in demosaicking performance. However, they are computationally intensive and rely on large model architectures, rendering them unsuitable for edge-devices deployment. This work introduces a novel demosaicking method to address these challenges based on green learning (GL), named green image demosaicking (GID). GID offers model transparency while significantly reducing the model size and computational complexity against deep learning methods. Notably, GID does not utilize neural networks. Instead, it is built upon unsupervised representation learning and supervised feature dimension reduction. GID effectively addresses the challenges of big data in vision applications and enhances predictive accuracy during decision-making. GID is engineered for rapid execution with parallel training, making it well-suited for real-time vision tasks on resource-constrained devices.

In Video-based Facial Expression Recognition (V-FER), models are typically trained on closed-set datasets with a fixed number of known classes. However, these models struggle with unknown classes common in real-world scenarios. In this paper, we introduce a challenging Open-set Video-based Facial Expression Recognition (OV-FER) task, aiming to identify both known and new, unseen facial expressions. While existing approaches use large-scale vision-language models like CLIP to identify unseen classes, we argue that these methods may not adequately capture the subtle human expressions needed for OV-FER. To address this limitation, we propose a novel Human Expression-Sensitive Prompting (HESP) mechanism to significantly enhance CLIP's ability to model video-based facial expression details effectively. Our proposed HESP comprises three components: 1) a textual prompting module with learnable prompts to enhance CLIP's textual representation of both known and unknown emotions, 2) a visual prompting module that encodes temporal emotional information from video frames using expression-sensitive attention, equipping CLIP with a new visual modeling ability to extract emotion-rich information, and 3) an open-set multi-task learning scheme that promotes interaction between the textual and visual modules, improving the understanding of novel human emotions in video sequences. Extensive experiments conducted on four OV-FER task settings demonstrate that HESP can significantly boost CLIP's performance (a relative improvement of 17.93% on AUROC and 106.18% on OSCR) and outperform other state-of-the-art open-set video understanding methods by a large margin. Code is available at https://github.com/cosinehuang/HESP.

Multiple unmanned aerial vehicles (UAVs) computation offloading plays a major role in the sixth generation (6G) mobile networks for resolving conflicts between computation-intensive or time-sensitive tasks and sensor nodes (SNs) with limited capabilities. Emerging artificial intelligence businesses such as multimodal data confusion and multi-video streams processing bring new challenges to UAV-aided communication and computation networks due to their multi-point and streaming characteristics. This paper focuses on a multi-UAV aided multi-point streaming tasks scenario where the SNs generate subtasks comprising various workflows randomly in a service period. We model the sequential relationship between multi-point tasks and time constraints for different workflows and aim to minimize the average energy consumption of all SNs while satisfying different time constraints by jointly optimizing the UAV-SN association policy and UAV deployment locations. To address the non-convexity problem, we decompose the problem into two subproblems, which we then solve alternately with the Markov approximation algorithm and convex optimization until convergence. Numerical results show that the proposed algorithm can effectively lower the average energy consumption of all SNs by a maximum of 35.6% compared to other schemes while maintaining a successful service ratio of 93%.

Visual tracking is a critical task in many computer vision applications such as surveillance and robotics. However, although the robustness to local corruptions has been improved, prevailing trackers are still sensitive to large scale corruptions, such as occlusions and illumination variations. In this paper, we propose a novel robust object tracking technique depends on subspace learning-based appearance model. Our contributions are twofold. First, mask templates produced by frame difference are introduced into our template dictionary. Since the mask templates contain abundant structure information of corruptions, the model could encode information about the corruptions on the object more efficiently. Meanwhile, the robustness of the tracker is further enhanced by adopting system dynamic, which considers the moving tendency of the object. Second, we provide the theoretic guarantee that by adapting the modulated template dictionary system, our new sparse model can be solved by the accelerated proximal gradient algorithm as efficient as in traditional sparse tracking methods. Extensive experimental evaluations demonstrate that our method significantly outperforms 21 other cutting-edge algorithms in both speed and tracking accuracy, especially when there are challenges such as pose variation, occlusion, and illumination changes.

An energy-efficient and reliable routing optimization scheme for mobile ad hoc networks (MANETs) is investigated in this paper. Three elements are regarded as the most important keys that affect system performance, these three elements are transmission power, interference, and link lifetime. The contribution of this paper is the development of an optimal routing algorithm called the Energy-Efficient and Reliable Routing (EERR). The EERR algorithm exploits location information to allocate optimal transmission power subject to the key elements above and adapts it for different network scenarios. The interaction of these elements and their effects on routing are studied. Simulations demonstrate that EERR has high network performance and outperforms PEER and AODV, a state-of-the-art.