Research map / 3D vision / Human pose estimation
Reconstructing Three-Dimensional Models of Interacting Humans
Mihai Fieraru, Mihai Zanfir, Elisabeta Oneaţă, Alin-Ionut Popa, Vlad Olaru, Cristian Sminchisescu
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 semantic 2d features as input (ablation) | ISP · FlickrCI3D (75 regions) | segmentation IoU: 0.300 -> 0.318 | Improved |
| add semantic 2d features as input (ablation) | ISP · FlickrCI3D (17 regions) | signature IoU: 0.240 -> 0.248 | Improved |
| add semantic 2d features as input (ablation) | ISP · FlickrCI3D (9 regions) | signature IoU: 0.410 -> 0.408 | No clear effect |
| add joint learning of contact segmentation (ablation) | ISP · FlickrCI3D (17 regions) | signature IoU: 0.218 -> 0.248 | Improved |
| add joint learning of contact segmentation (ablation) | ISP · FlickrCI3D (75 regions) | signature IoU: 0.072 -> 0.082 | Improved |
| add masking of correspondences outside estimated segmentation (ablation) | ISP · FlickrCI3D (17 regions) | signature IoU: 0.230 -> 0.248 | Improved |
| add masking of correspondences outside estimated segmentation (ablation) | ISP · FlickrCI3D (9 regions) | signature IoU: 0.385 -> 0.408 | Improved |
| coarser region granularity (75 to 9 regions) (ablation) | ISP · FlickrCI3D | signature IoU: 0.082 -> 0.408 | Improved |
| add geometric alignment term L_G on contact signatures (ablation) | optimization L* · CHI3D contact frames (overall) | pose error (mm): 130.7 (w/o L_G) vs Hug 180.4 -> 173.9 | Improved |
| add geometric alignment term L_G on contact signatures (ablation) | optimization L* · CHI3D contact frames (overall) | translation error (mm): 408.4 mm without L_G; Kick 460.1 -> 419.2 | Improved |
| add geometric alignment term L_G on contact signatures (ablation) | optimization L* · CHI3D Holding hands | pose error (mm): 100.3 -> 100.9 | Hurt |
| add geometric alignment term L_G on contact signatures (ablation) | optimization L* · CHI3D contact frames (overall) | mean contact distance (mm): 420.8 -> 26.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.
- Cliff — CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimation (ECCV 2022) · better on 0 of 1 shared result
- REMIPS — REMIPS: Physically Consistent 3D Reconstruction of Multiple Interacting People under Weak Supervision (NeurIPS 2021) · better on 0 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
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · similarity 0.91)
- ContactField: Implicit Field Representation for Multi-Person Interaction Geometry (NeurIPS 2024 · similarity 0.90)
- Pose Priors from Language Models (CVPR 2025 · similarity 0.90)
- InteractVLM: 3D Interaction Reasoning from 2D Foundational Models (CVPR 2025 · similarity 0.90)
- PICO: Reconstructing 3D People In Contact with Objects (CVPR 2025 · similarity 0.89)
- HOI-M^3: Capture Multiple Humans and Objects Interaction within Contextual Environment (CVPR 2024 · similarity 0.88)
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.
