Multi-Person 3D HumanPose Dataset (MP-3DHP) is a depth sensor-based dataset, which was constructed to facilitate the development of multi-person 3D pose estimation methods targeting real-world challenges. The dataset includes 177k training data and 33k validation data where both the 3D human poses and body segments are avaliable. The dataset also include 9k clean background data and 4k testing data including multi-person 3D poses.
Paper | Code | Results | Date | Stars |
---|