Research map / Label-efficient and robust learning / Semi-supervised learning
ScarceNet: Animal Pose Estimation With Scarce Annotations
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 |
|---|---|---|---|
| Use ScarceNet instead of FlexMatch | AP-10K semi-supervised pose estimation · AP-10K, 5 images/species | mAP: 0.466 -> 0.533 | Helped |
| Use ScarceNet instead of FlexMatch | AP-10K semi-supervised pose estimation · AP-10K, 25 images/species | mAP: 0.646 -> 0.681 | Helped |
| Add reusable sample re-labeling (ablation) | ScarceNet · AP-10K, 5 images/species | mAP: 0.521 -> 0.533 | Helped |
| Add student-teacher consistency (ablation) | ScarceNet · AP-10K, 5 images/species | mAP: 0.487 -> 0.533 | Helped |
| Add multi-branch architecture (ablation) | ScarceNet · AP-10K, 5 images/species | mAP: 0.520 -> 0.533 | Helped |
| Use strong-weak augmentation instead of common rotation and scale augmentation (ablation) | ScarceNet · AP-10K, 5 images/species | mAP: 0.487 -> 0.533 | Helped |
| Add reusable sample re-labeling (ablation) | ScarceNet · AP-10K, 25 images/species | mAP: 0.668 -> 0.681 | Helped |
| Add student-teacher consistency (ablation) | ScarceNet · AP-10K, 25 images/species | mAP: 0.659 -> 0.681 | Helped |
| Add multi-branch architecture (ablation) | ScarceNet · AP-10K, 25 images/species | mAP: 0.670 -> 0.681 | Helped |
| Use strong-weak augmentation instead of common rotation and scale augmentation (ablation) | ScarceNet · AP-10K, 25 images/species | mAP: 0.654 -> 0.681 | Helped |
| Train with scarce labeled data and unlabeled data using ScarceNet instead of full supervision with synthetic-data domain adaptation | TigDog · TigDog average | PCK@0.05: MDAM-MT 73.66; Ours 73.83; Real 80.48 | Mixed |
Closest papers by meaning
- From Synthetic to Real: Unsupervised Domain Adaptation for Animal Pose Estimation (CVPR 2021 · similarity 0.92)
- AP-10K: A Benchmark for Animal Pose Estimation in the Wild (NeurIPS 2021 · similarity 0.90)
- APT-36K: A Large-scale Benchmark for Animal Pose Estimation and Tracking (NeurIPS 2022 · similarity 0.89)
- Probabilistic Prompt Distribution Learning for Animal Pose Estimation (CVPR 2025 · similarity 0.88)
- Pose Splatter: A 3D Gaussian Splatting Model for Quantifying Animal Pose and Appearance (NeurIPS 2025 · similarity 0.87)
- UniAP: Towards Universal Animal Perception in Vision via Few-Shot Learning (AAAI 2024 · similarity 0.84)
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.