ASR-GCN: Adaptive spatial information reconstruction GCN for skeleton-based action recognition.

Cyclopropyl as a versatile tool in the development of kinase-targeted therapeutics.

FMFNet: A Faster Multimodal Fusion Network for action recognition via efficient modality compensation

Continuous Learning Enabled Dual Attention-based Fairness-aware eXplainable Convolutional Network for bias mitigation in a critical decision system for human action recognition

TE-STGCN: Topology enhanced spatio-temporal graph convolutional network for skeleton-based action recognition

Multimodal alignment of event and text streams in spiking neural networks for human action recognition

Domain adaptation and generalization are crucial for real-world applications, such as autonomous driving and medical imaging where the model must operate reliably across environments with distinct data distributions. However, these tasks are challenging because the model needs to overcome various domain gaps caused by variations in, for example, lighting, weather, sensor configurations, and so on. Addressing domain gaps simultaneously in different modalities, known as multimodal domain adaptation and generalization, is even more challenging due to unique challenges in different modalities. Over the past few years, significant progress has been made in these areas, with applications ranging from action recognition to semantic segmentation, and more. Recently, the emergence of large-scale pre-trained multimodal foundation models, such as CLIP, has inspired numerous research studies, which leverage these models to enhance downstream adaptation and generalization. This survey summarizes recent advances in multimodal adaptation and generalization, particularly how these areas evolve from traditional approaches to foundation models. Specifically, this survey covers (1) multimodal domain adaptation, (2) multimodal test-time adaptation, (3) multimodal domain generalization, (4) domain adaptation and generalization with the help of multimodal foundation models, and (5) adaptation of multimodal foundation models. For each topic, we formally define the problem and give a thorough review of existing methods. Additionally, we analyze relevant datasets and applications, highlighting open challenges and potential future research directions.

A data-centric approach to radar-based human action recognition: SVD-based clutter removal and RTM/DTM feature fusion

Motion-aware multi-level feature fusion for few-shot action recognition

Hierarchical joint contrastive learning with knowledge distillation for self-supervised 3D skeleton-based action recognition

An unsupervised open-set recognition method for user-independent human activity recognition

A systematic review of deep learning-based models for elderly and human activity recognition

ESP-Fi HAR: A low-power WiFi CSI dataset for Ad-Hoc IoT human activity recognition

Pedestrian navigation activity recognition method based on two-stream transformer and contrastive learning.

Adaptive temporal convolutional network with multi-head EMA-gated attention for continuous radar-based human activity recognition

Benchmarking Radar Preprocessing Techniques and Transfer Learning Models for FMCW-based Human Activity Recognition

Ensuring occupational safety in shipyard welding demands reliable activity recognition under constrained postures, mechanical vibration, and electromagnetic interference.This study presents a wearable smartphone-based Welding Activity Recognition (WAR) framework employing 9-DoF inertial sensing and a deployment-oriented evaluation protocol.Triaxial accelerometer, gyroscope, and magnetometer signals from ten certified welders were processed using Kalman filtering and 2-second sliding-window segmentation with 50% overlap under a subjectindependent scheme.Five classifiers, SVM, LightGBM, CNN, LSTM, and a hybrid CNN-LSTM were comparatively evaluated.The proposed CNN-LSTM achieved the highest accuracy of 97.45% at 50 Hz, demonstrating effective spatio-temporal feature modeling for multi-class welding posture recognition.Latency was rigorously defined under a unified sensor-to-decision framework, separating buffering and computational components.Despite lower inference latency in classical models, the proposed architecture maintained computational latency within the steady-state streaming interval, ensuring real-time feasibility.Additional analyses of sensor placement, sampling rate, and magnetometer inclusion provide further substantiate robust deployment in industrial environments.

Human Activity Recognition (HAR) is emerging as a critical enabler of context-aware applications in healthcare, fitness, and smart environments.In this research we present an approach that involves the Hierarchical Fusion with Decision Enhancement for Human Activity Recognition (Hi-FuseDE-HAR) framework.It contains four sequential hierarchical levels that transform raw sensor signals into a reliable HAR decision.At the input level, heterogeneous data streams are obtained from multiple wearable and ambient sensors.Applying Level 0 build the discriminative latent embedding for each sensor modality separately.Level 1 fuses across sensors in groupings and determines the value of using each modality and determines its relative contribution to the overall feature importance.Level 2 applies a Graph Cross-Modal Transformer that learns relationship between sensors groups producing a globally consistent fused representation.Level 3 provides decision enhancement through uncertainty calibration and utility aware optimization to ensure the final estimates based.Experimental results indicate that the proposed framework achieves 97.6% accuracy and 96.7% F1-score on the PAMAP2 dataset, 95.5% accuracy and 93.2% F1-score on the OPPORTUNITY dataset, and 96.5% accuracy and 95.2% F1-score on the MHEALTH dataset respectively.Notably, Hi-FuseDE-HAR retains strong performance confirming its capability to generalize across varied sensor contexts and complex activity patterns.

To address the problem of low recognition rate caused by the difficulty in capturing highspeed and subtle movements in table tennis, this work proposes a motion recognition method based on multimodal data and an optimized Spatial-Temporal Graph Convolutional Network (ST-GCN). The model introduces a Multi-Level Graph Convolutional Network (ML-GCN) architecture and constructs cross-level feature extraction channels, which effectively capture the spatiotemporal correlations between local subtle movements and global trajectories. The built-in hybrid attention mechanism realizes precise focusing on key skeletal nodes and core motion frames through adaptive weight assignment. Combined with the multimodal fusion strategy of visual signals and inertial sensor data, it significantly enhances the robustness of the model in scenarios with line-of-sight occlusion and motion blur. Test results based on a self-built multimodal table tennis dataset show that this method achieves an accuracy of 88.2%, a recall rate of 89.5% and an F1-score of 88.3%. This performance is significantly superior to the original ST-GCN and existing mainstream motion recognition algorithms, which confirms the core role of each optimization module in improving feature representation capability and computational efficiency. The study provides an efficient technical solution for the intelligent analysis of complex sports movements.

In college physical education teaching, college students’ athletic ability assessment still faces problems of strong subjectivity in scoring and difficulty in quantifying complex movement characteristics. Traditional methods struggle to capture multi-granularity information of skeletons in continuous movements, and the cost of acquiring annotated data is high. This study aims to construct a set of quantitative assessment models for athletic ability oriented to college physical education classrooms. It realizes the joint optimization of action recognition and athletic ability scoring through the multi-granularity spatial-temporal graph convolutional network (MGSTGC). The MG-STGC model uses an encoder to extract joint-level, limb-level, and body-level features. It combines labeled and unlabeled data via semi-supervised learning strategies to achieve the joint optimization of action recognition and quantitative assessment of athletic ability. The athletic ability assessment module can generate continuous scores across four dimensions: strength, stability, standardization, and coordination. These scores are obtained through spatiotemporal statistical mapping of historical action segments and skeleton features, providing data support for individualized training. On the NTU RGB+D dataset, MG-STGC achieves a Top-1 accuracy of 95.6% and 89.7% on the X-view and X-sub benchmarks. On the FineGym dataset, it reaches 80.5% Top-1 accuracy on the Gym99 subset and 75.4% on the Gym288 subset, with category average accuracies of 69.8% and 62.6%, respectively, outperforming baseline models. Ablation experiments show that the granularity information fusion module and parameters have an impact on model performance. Research shows that MGSTGC can efficiently capture multi-granularity information of action skeletons and provide an objective and quantitative method for athletic ability assessment in college physical education classrooms. MG-STGC also lays a feasible theoretical and practical foundation for intelligent physical education teaching and personalized training.