Multimodal Deep Learning for Robust Human Behavior Recognition
This research proposes a multimodal deep learning approach that effectively integrates visual and audio data to achieve highly accurate human behavior recognition, outperforming unimodal techniques.