CLIP-MG: Guiding Semantic Attention with Skeletal Pose Features and RGB Data for Micro-Gesture Recognition on the iMiGUE Dataset

Open in new window