HumanML-3D
Emerging6papers using it
2023first seen
HumanML-3D is a dataset used to evaluate the generation of 3D human motion based on textual descriptions, focusing on the diversity, naturalness, and adherence of the generated motion to the provided text.
Papers using HumanML-3D (6)
- Cross-Modal Retrieval for Motion and Text via DropTriple LossMotionVLA: Vision-Language-Action Model for Humanoid MotionFine-grained Motion Retrieval via Joint-Angle Motion Images and Token-Patch Late InteractionMulti-Modal Motion Retrieval by Learning a Fine-Grained Joint Embedding SpaceTri-Modal Motion Retrieval by Learning a Joint Embedding SpaceAttT2M: Text-Driven Human Motion Generation with Multi-Perspective
Attention Mechanism