Existing multi-person video pose estimation methods typically adopt a two-stage pipeline: detecting individuals in each frame, followed by temporal modeling for single-person pose estimation. This design relies on heuristic operations such as tracking, RoI cropping, and non-maximum suppression, limiting both accuracy and efficiency. In this paper, we present a fully end-to-end framework for multi-person 2D pose estimation in videos, effectively eliminating heuristic operations. A key challenge is to associate individuals across frames under complex and overlapping temporal trajectories. To address this, we introduce a novel Pose-Aware Video transformEr Network (PAVE-Net), which features a spatial encoder to model intra-frame relations and a spatiotemporal pose decoder to capture global dependencies across frames. To achieve accurate temporal association, we propose a pose-aware attention mechanism that enables each pose query to selectively aggregate features corresponding to the same individual across consecutive frames. Additionally, we explicitly model spatiotemporal dependencies among pose keypoints to improve accuracy. Notably, our approach is the first end-to-end method for multi-frame 2D human pose estimation. Extensive experiments show that PAVE-Net substantially outperforms prior image-based end-to-end methods, achieving a 6.0 mAP improvement on PoseTrack2017, and delivers accuracy competitive with state-of-the-art two-stage video-based approaches, while offering significant gains in efficiency.

Academic emotion analysis plays a crucial role in evaluating students' engagement and cognitive states during the learning process. This paper addresses the challenge of automatically recognizing academic emotions through facial expressions in real-world learning environments. While significant progress has been made in facial expression recognition for basic emotions, academic emotion recognition remains underexplored, largely due to the scarcity of publicly available datasets. To bridge this gap, we introduce RAER, a novel dataset comprising approximately 2,700 video clips collected from around 140 students in diverse, natural learning contexts such as classrooms, libraries, laboratories, and dormitories, covering both classroom sessions and individual study. Each clip was annotated independently by approximately ten annotators using two distinct sets of academic emotion labels with varying granularity, enhancing annotation consistency and reliability. To our knowledge, RAER is the first dataset capturing diverse natural learning scenarios. Observing that annotators naturally consider context cues-such as whether a student is looking at a phone or reading a book-alongside facial expressions, we propose CLIP-CAER (CLIP-based Context-aware Academic Emotion Recognition). Our method utilizes learnable text prompts within the vision-language model CLIP to effectively integrate facial expression and context cues from videos. Experimental results demonstrate that CLIP-CAER substantially outperforms state-of-the-art video-based facial expression recognition methods, which are primarily designed for basic emotions, emphasizing the crucial role of context in accurately recognizing academic emotions. Project page: https://zgsfer.github.io/CAER

Due to the flourishing of Internet of Things (IoT) technology, radar-based human activity recognition (HAR) technology has made significant progress and has become an indispensable research area. Many radar systems use multiple feature maps and handle them directly in image format. However, the generation of multiple forms of feature maps requires heavy computational resources, which makes it impractical for real-world applications. Additionally, many networks fail to fully extract temporal information from the time-Doppler (TD) map during the feature extraction. Therefore, a time-weighted network based on strip pooling (TWN-SP) using the TD map is proposed in this article. The TWN-SP consists of two time-weighted modules based on fire (TWMs-F) and a feature fusion module based on temporal and channel attention (FFM-TCA). Due to the application of depthwise separable convolution (DSC) and placing the feature fusion step at the forefront, the proposed network has fewer parameters. Moreover, a radar dataset named RadSet is constructed, containing TD maps of six daily activities. To validate the performance of the TWN-SP, a ten-fold cross-validation (CV) on the public dataset named Radar848 and a leave-one-subject-out (LOSO) CV on the RadSet dataset are carried out, respectively. To further validate the generalization performance of TWN-SP, a comparative analysis was conducted on another public dataset, Ci4R. The TWN-SP achieves an accuracy of 99.28% on the RadSet dataset. Experimental results indicate that the TWN-SP surpasses the current leading networks in both performance and complexity.