Research map / 3D vision / Human pose estimation
Generative Zoo
Tomasz Niewiadomski, Anastasios Yiannakidis, Hanz Cuevas-Velasquez, Soubhik Sanyal, Michael J. Black, Silvia Zuffi, Peter Kulits
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 depth control to ControlNet conditioning (ablation) | GenZoo pipeline · Animal3D (100k samples) | S-MPJPE: 184.1 -> 166.9 | Improved |
| add depth control to ControlNet conditioning (ablation) | GenZoo pipeline · GenZoo-Felidae (100k samples) | S-V2V: 95.4 -> 59.3 | Improved |
| add Canny-edge control alongside depth (ablation) | GenZoo pipeline · Animal3D (100k samples) | S-MPJPE: 172.3 -> 166.9 | Improved |
| add Canny-edge control alongside depth (ablation) | GenZoo pipeline · GenZoo-Felidae (100k samples) | PA-V2V: 39.1 -> 50.2 | Hurt |
| add VLM orientation captioning to the prompt (ablation) | GenZoo pipeline · GenZoo-Felidae (100k samples) | S-V2V: 71.0 -> 59.3 | Improved |
| add VLM orientation captioning to the prompt (ablation) | GenZoo pipeline · Animal3D (100k samples) | PA-MPJPE: 120.1 -> 118.4 | Improved |
| add LLM prompt synthesis (ablation) | GenZoo pipeline · GenZoo-Felidae (100k samples) | S-V2V: 69.4 -> 59.3 | Improved |
| add LLM prompt synthesis (ablation) | GenZoo pipeline · Animal3D (100k samples) | PCK@0.5: 97.2 -> 97.1 | No clear effect |
| use FLUX instead of Hunyuan-DiT as image generator (ablation) | GenZoo pipeline · Animal3D (100k samples) | S-MPJPE: 174.0 -> 166.9 | Improved |
| use FLUX instead of Stable Diffusion 3 as image generator (ablation) | GenZoo pipeline · Animal3D (100k samples) | S-MPJPE: 178.3 -> 166.9 | Improved |
| use FLUX instead of Stable Diffusion 3 as image generator (ablation) | GenZoo pipeline · Animal3D (100k samples) | PCK@0.5: 97.5 -> 97.1 | Hurt |
| use FLUX instead of Hunyuan-DiT as image generator (ablation) | GenZoo pipeline · GenZoo-Felidae (100k samples) | PA-V2V: 47.6 -> 50.2 | Hurt |
Closest papers by meaning
- Animal3D: A Comprehensive Dataset of 3D Animal Pose and Shape (ICCV 2023 · similarity 0.94)
- Pose Splatter: A 3D Gaussian Splatting Model for Quantifying Animal Pose and Appearance (NeurIPS 2025 · similarity 0.89)
- SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild (NeurIPS 2026 · similarity 0.89)
- Learning the 3D Fauna of the Web (CVPR 2024 · similarity 0.88)
- BEAST3D: animal behavioral analysis and neural encoding from multi-view video via Gaussian splatting (NeurIPS 2026 · similarity 0.88)
- Web-Scale Collection of Video Data for 4D Animal Reconstruction (NeurIPS 2025 · 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.
