AI learns which body parts to edit when given text instructions

Original: Cross-Axis Feature Fusion with Joint-Wise Motion Difference Prediction for Text-Based 3D Human Motion Editing

Source: arxiv.org ↗