Research map / 3D vision / Human pose estimation
ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Video
Xiaozhong Lyu, Gen Li, Zhiyin Qian, Xucong Zhang, Marc Pollefeys, Siyu Tang
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 joint multimodal training instead of a body-only RGB specialist (ablation) | ReViV · ADT egocentric body motion | GA-MPJPE: 200.3 -> 111.5 | Improved |
| add large datasets without body annotations to training (ablation) | ReViV · ADT egocentric body motion | GA-MPJPE: 134.0 -> 111.5 | Improved |
| use the Dirichlet-based masking sampling instead of 1-to-1 masking (ablation) | ReViV · ADT egocentric body motion | GA-MPJPE: 158.3 -> 111.5 | Improved |
| use the proposed mask sampling instead of uniform sampling over modalities (ablation) | ReViV · ADT egocentric body motion | GA-MPJPE: 172.7 -> 111.5 | Improved |
| scale model from ~127M Tiny to full size (ablation) | ReViV · ADT egocentric body motion | GA-MPJPE: 143.1 -> 111.5 | Improved |
| add Nymeria to training data (ablation) | ReViV · ADT egocentric depth | delta<1.25: 32.0 -> 56.5 | Improved |
| add HoloAssist to training data (with Nymeria) (ablation) | ReViV · ADT egocentric depth | delta<1.25: 30.1 -> 32.0 | Improved |
| add continuous ViT encoding branch (ablation) | ReViV · ADT egocentric depth | Abs Rel: 0.296 -> 0.265 | Improved |
| add continuous ViT encoding branch (ablation) | ReViV · ADT egocentric camera tracking | RRE: 1.258 -> 1.279 | Hurt |
| use joint training instead of hand specialist (ablation) | ReViV · TACO hand | GA-MPJPE: 53.2 -> 20.3 | Improved |
| use ReViV instead of Dyn-HaMR for hand reconstruction | ReViV · HOT3D hand | GA-MPJPE: 107.7 -> 38.2 | Improved |
| use ReViV (no camera) instead of UniEgoMotion with GT camera | ReViV · ADT egocentric body motion | GA-MPJPE: 105.8 -> 111.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Dyn-HaMR — Dyn-HaMR: Recovering 4D Interacting Hand Motion from a Dynamic Camera (CVPR 2025) · better on 3 of 3 shared results
- EgoM2P — EgoM2P: Egocentric Multimodal Multitask Pretraining (ICCV 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.88)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · similarity 0.88)
- Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives (CVPR 2024 · similarity 0.87)
- UniCon3R: Unified Contact-aware 4D Human-Scene Reconstruction from Monocular Video (NeurIPS 2026 · similarity 0.87)
- 3D Human Pose Perception from Egocentric Stereo Videos (CVPR 2024 · similarity 0.87)
- EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Videos Generation (NeurIPS 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.
