Research map / 3D vision / Human pose estimation
AP-10K: A Benchmark for Animal Pose Estimation in the Wild
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 ImageNet pretraining (ablation) | HRNet-w32 · AP-10K val | mAP: 0.703 -> 0.738 | Helped |
| add ImageNet pretraining (ablation) | ResNet50 · AP-10K val | mAP: 0.646 -> 0.699 | Helped |
| train from scratch 3x longer (630 epochs) vs ImageNet pretraining (ablation) | HRNet-w32 · AP-10K val | mAP: 0.725 vs 0.735 | No clear effect |
| longer schedule for ImageNet-pretrained model (ablation) | HRNet-w32 · AP-10K val | mAP: 0.738 -> 0.735 | No clear effect |
| COCO human-pose pretraining with short finetune (20 epochs) (ablation) | HRNet-w32 · AP-10K val | mAP: 0.738 -> 0.606 | Hurt |
| COCO human-pose pretraining with 210-epoch finetune instead of ImageNet pretraining (ablation) | HRNet-w32 · AP-10K val | mAP: 0.738 -> 0.753 | Helped |
| train on all species instead of same-family other species (ablation) | HRNet-w32 · Sheep | mAP: 0.663 -> 0.761 | Helped |
| train on Bovidae only, test on distant family (Hominidae) (ablation) | HRNet-w32 · Hominidae | mAP: 0.782 -> 0.015 | Hurt |
| 20-shot finetuning for unseen species (ablation) | HRNet-w32 · Zebra | mAP: 0.324 -> 0.480 | Helped |
| finetune on 140 images (ablation) | HRNet-w32 · Chimpanzee | mAP: 0.022 -> 0.550 | Helped |
| train on AP-10K vs Animal-Pose dataset for cross-dataset transfer | HRNet-w32 · Animal-Pose test | mAP: 0.424 -> 0.913 | Helped |
| use HRNet instead of ResNet/Hourglass | HRNet-w32 · AP-10K val | mAP: 0.699 -> 0.738 | Helped |
Closest papers by meaning
- Probabilistic Prompt Distribution Learning for Animal Pose Estimation (CVPR 2025 · similarity 0.89)
- Pose Splatter: A 3D Gaussian Splatting Model for Quantifying Animal Pose and Appearance (NeurIPS 2025 · similarity 0.88)
- Animal-Bench: Benchmarking Multimodal Video Models for Animal-centric Video Understanding (NeurIPS 2024 · similarity 0.88)
- Web-Scale Collection of Video Data for 4D Animal Reconstruction (NeurIPS 2025 · similarity 0.87)
- UniAP: Towards Universal Animal Perception in Vision via Few-Shot Learning (AAAI 2024 · similarity 0.87)
- Reconstructing Animals and the Wild (CVPR 2025 · similarity 0.86)
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.