Research map / 3D vision / Human pose estimation
UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture
Hiroyasu Akada, Jian Wang, Soshi Shimada, Masaki Takahashi, Christian Theobalt, Vladislav Golyanik
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use a stereo 2D module with shared encoders and joint decoder instead of EgoGlass separate UNets | egocentric 3D pose estimator · UnrealEgo test | MPJPE: 91.44 -> 79.06 | Improved |
| use stereo input instead of monocular xR-EgoPose | egocentric 3D pose estimator · UnrealEgo test | MPJPE: 112.86 -> 79.06 | Improved |
| add ImageNet pre-training of the encoder | Ours · UnrealEgo test | MPJPE: 87.31 -> 79.06 | Improved |
| add ImageNet pre-training of the encoder | Ours · UnrealEgo test | PA-MPJPE: 64.65 -> 59.95 | Improved |
| add weight sharing between the two stereo encoders (ablation) | Ours 2D module · UnrealEgo test | MPJPE: 83.54 -> 79.06 | Improved |
| add weight sharing between the two stereo encoders (ablation) | Ours 2D module · UnrealEgo test | PA-MPJPE: 62.29 -> 59.95 | Improved |
| use separate training of 2D and 3D modules instead of end-to-end (ablation) | Ours · UnrealEgo test | MPJPE: 80.67 -> 79.06 | Improved |
| use separate training of 2D and 3D modules instead of end-to-end (ablation) | Ours · UnrealEgo test | PA-MPJPE: 61.72 -> 59.95 | Improved |
| use larger encoder backbone ResNet50 instead of ResNet18 (ablation) | Ours 2D module · UnrealEgo test | PA-MPJPE: 59.95 -> 60.08 | No clear effect |
| use larger encoder backbone ResNet101 instead of ResNet18 (ablation) | Ours 2D module · UnrealEgo test | MPJPE: 79.06 -> 80.15 | Hurt |
Closest papers by meaning
- 3D Human Pose Perception from Egocentric Stereo Videos (CVPR 2024 · similarity 0.93)
- EventEgo3D: 3D Human Motion Capture from Egocentric Event Streams (CVPR 2024 · similarity 0.88)
- FreeMan: Towards Benchmarking 3D Human Pose Estimation under Real-World Conditions (CVPR 2024 · similarity 0.88)
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.88)
- HOT3D: Hand and Object Tracking in 3D from Egocentric Multi-View Videos (CVPR 2025 · similarity 0.87)
- FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
