Research map / 3D vision / Face and shape
Animal Avatars: Reconstructing Animatable 3D Animals from Casual Videos
Remy Sabathier, David Novotny, Niloy Mitra
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 |
|---|---|---|---|
| factor camera motion out using fixed SfM camera instead of root pose representing both (ablation) | Animal Avatars · CoP3D | PSNR: 19.12 -> 22.40 | Improved |
| add Chamfer mask loss (ablation) | Animal Avatars · CoP3D | IoU: 0.70 -> 0.84 | Improved |
| add CSE keypoint loss (ablation) | Animal Avatars · CoP3D | PSNR: 20.89 -> 22.40 | Improved |
| add sparse keypoint loss alongside CSE (ablation) | Animal Avatars · CoP3D | PSNR: 21.54 -> 22.40 | Improved |
| add edge-length regularizer (ablation) | Animal Avatars · CoP3D | IoU: 0.83 -> 0.84 | Improved |
| use video-based CSE-guided template fitting instead of RAC | Animal Avatars · CoP3D 15/5 | LPIPS: 0.164 -> 0.041 | Improved |
| use temporally consistent video fitting instead of per-frame BITE | Animal Avatars · CoP3D | IoUw5: 0.59 -> 0.79 | Improved |
| use Animal Avatars instead of BITE | CoP3D mean IoU · CoP3D | IoU: 0.81 -> 0.84 | Improved |
| reduce supervision with larger test blocks (15/5 to 15/15 split) (ablation) | Animal Avatars · CoP3D | IoU: 0.84 -> 0.81 | 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.
- BITE — BITE: Beyond Priors for Improved Three-D Dog Pose Estimation (CVPR 2023) · better on 3 of 3 shared results
- BARC — BARC: Learning To Regress 3D Dog Shape From Images by Exploiting Breed Information (CVPR 2022) · better on 3 of 3 shared results
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
- Who Left the Dogs Out? 3D Animal Reconstruction with Expectation Maximization in the Loop (ECCV 2020 · similarity 0.89)
- Coarse-to-fine Animal Pose and Shape Estimation (NeurIPS 2021 · similarity 0.89)
- Progressive Pose-Guided 4D Animal Reconstruction from Monocular Video (ECCV 2026 · similarity 0.89)
- DogRecon: Canine Prior-Guided Animatable 3D Gaussian Dog Reconstruction From A Single Image (IJCV 2025 · similarity 0.89)
- SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild (NeurIPS 2026 · similarity 0.87)
- SAOR: Single-View Articulated Object Reconstruction (CVPR 2024 · 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.
