Fetching the paper…

AttT2M: Text-Driven Human Motion Generation with Multi-Perspective Attention Mechanism · Around