Research map / 3D vision / Human pose estimation
Through-Wall Human Mesh Recovery Using Radio Signals
Mingmin Zhao, Yingcheng Liu, Aniruddh Raghu, Tianhong Li, Hang Zhao, Antonio Torralba, Dina Katabi
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 |
|---|---|---|---|
| add the temporal pose and dynamics discriminator instead of a frame-based discriminator (ablation) | RF-Avatar · RF dataset | MPJPE (cm): 6.88 -> 6.05 | Improved |
| add the temporal pose and dynamics discriminator instead of a frame-based discriminator (ablation) | RF-Avatar · RF dataset | Per-vertex error (cm): 2.24 -> 1.88 | Improved |
| add strong supervision on shape vector beta (ablation) | RF-Avatar · RF dataset | Per-vertex error (cm): 4.70 -> 1.88 | Improved |
| add strong supervision on shape vector beta (ablation) | RF-Avatar · RF dataset | MPJPE (cm): 6.72 -> 6.05 | Improved |
| add multi-headed temporal self-attention for shape (ablation) | RF-Avatar · RF dataset | Per-vertex error (cm): 2.55 -> 1.88 | Improved |
| add multi-headed temporal self-attention for shape (ablation) | RF-Avatar · RF dataset | MPJPE (cm): 6.43 -> 6.05 | Improved |
| use ground truth boxes instead of TPN-detected trajectories | RF-Avatar · RF dataset | MPJPE (cm): 6.05 -> 5.75 | No clear effect |
| evaluate through a wall rather than line-of-sight, without training on through-wall data | RF-Avatar · Through-wall vs line-of-sight | MPJPE (cm): 5.84 -> 6.26 | No clear effect |
Closest papers by meaning
- Through-Wall Human Pose Estimation Using Radio Signals (CVPR 2018 · similarity 0.88)
- Human Mesh Recovery From Monocular Images via a Skeleton-Disentangled Representation (ICCV 2019 · similarity 0.85)
- Learning Human Mesh Recovery in 3D Scenes (CVPR 2023 · similarity 0.84)
- GLAMR: Global Occlusion-Aware Human Mesh Recovery With Dynamic Cameras (CVPR 2022 · similarity 0.84)
- Pose Modulated Avatars from Video (ICLR 2024 · similarity 0.83)
- DPMesh: Exploiting Diffusion Prior for Occluded Human Mesh Recovery (CVPR 2024 · similarity 0.83)
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.
