Text-driven human motion generation surveys, datasets and models, focusing on single-person scenarios without human-object/scene interaction. Please refer to the project page for a better browsing ...