International Journal of Computational Intelligence Systems · 2023 · 12 citations · 43 references
ComputingVision ScienceComputer VisionSensor FusionVisual AttentionAction RecognitionDeep LearningFeature DescriptionVideo AnalysisActivity RecognitionAttention-based Action RecognitionFeature PoolingFeature Pooling ModelsHuman ActionSkeleton-based Action RecognitionSpatiotemporal Feature ExtractionTemporal Feature Extraction
Abstract In this paper, we propose VT-BPAN, a novel approach that combines the capabilities of Vision Transformer (VT), bilinear pooling, and attention network fusion for effective human action recognition (HAR). The proposed methodology significantly enhances the accuracy of activity recognition through the following advancements: (1) The introduction of an effective two-stream feature pooling and fusion mechanism that combines RGB frames and skeleton data to augment the spatial–temporal feature representation. (2) The development of a spatial lightweight vision transformer that mitigates computational costs. The evaluation of this framework encompasses three widely employed video action datasets, demonstrating that the proposed approach achieves performance on par with state-of-the-art methods.
43
DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · 2023 · 76.3K citations · Full text
ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks
Qilong Wang, Banggu Wu, Pengfei Zhu et al. · 2020 · 8.1K citations
Long-term recurrent convolutional networks for visual recognition and description
Jeff Donahue, Lisa Anne Hendricks, Sergio Guadarrama et al. · 2015 · 5.3K citations
Microsoft Kinect Sensor and Its Effect
Zhengyou Zhang · IEEE Multimedia · 2012 · 2.5K citations
A survey on vision-based human action recognition
Ronald Poppe · Image and Vision Computing · 2009 · 2.3K citations