Motion-X++: A Large-Scale Multimodal 3D Whole-Body Human Motion Dataset

Motion-X++ is a large-scale multimodal dataset for 3D whole-body human motion. It includes 2D keypoints for mesh recovery and motion generation, along with SMPL-X annotations differentiating translations and orientations in camera and world coordinate systems. The dataset, created by YuhongZhang, was last updated on January 14, -2025.

Use Cases

• Training models for 3D human mesh recovery based on provided 2D keypoints.
• Developing motion generation algorithms using the multimodal motion data.
• Analyzing human motion in different coordinate systems based on the SMPL-X annotations.
• Conducting research on action recognition or description using the included action descriptions.

Strengths

• Described as a large-scale dataset, suggesting substantial volume.
• Provides multimodal data including 2D keypoints and SMPL-X annotations.
• Includes action descriptions, adding a semantic layer to the motion data.

Limitations

• Row count, file formats, and exact size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• License information is unknown, which could restrict usage.

Provenance

Source

huggingface, author YuhongZhang

Collection Method

Method of data gathering is not described.

Time Range

null

Freshness

Last updated 2025-01-14 11:31:54.

Geography

null