Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Active Learning for Animal Re-Identification with Ambiguity-Aware Sampling
Depanshu Sani, Mehar Khurana, Saket Anand
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 ambiguity-aware sampling (AAS) with annotation on top of SpCL | SpCL · 13 wildlife datasets | mAP: 44.95% -> 56.14% | Improved |
| Use AAS instead of LBAS pair sampling | SpCL + AL · 13 wildlife datasets | mAP: 52.15% -> 56.14% | Improved |
| Use AAS instead of NIS sampling | SpCL + AL · 13 wildlife datasets | AUC ROC: 72.30% -> 75.21% | Improved |
| Add random pair annotation to SpCL | SpCL · 13 wildlife datasets | mAP: 44.95% -> 46.67% | Improved |
| Use AAS with less budget (0.033%) instead of 0.1% | SpCL + AL · 13 wildlife datasets | mAP vs budget: AAS 0.033% budget gives 56.14% vs LBAS 0.1% 52.15% | Improved |
| Use foundation models instead of pretrained ResNet-50 | ResNet-50 baseline · 13 wildlife datasets | mAP: 38.88% -> 45.65% (MegaDescriptor); 41.16% (MiewID) | Improved |
| Use AAS instead of LBAS on Market-1501 | SpCL + AL · Market-1501 | mAP: 67.38% -> 69.33% | Improved |
| Use AAS instead of LBAS on Person-X | SpCL + AL · Person-X | mAP: 60.48% -> 64.25% | Improved |
| Add random pair annotation to SpCL | SpCL · Market-1501 | mAP: 66.23% -> 65.10% | Hurt |
| Use NIS sampling instead of base SpCL | SpCL · Market-1501 | mAP: 66.23% -> 66.05% | No clear effect |
Closest papers by meaning
- Species-Agnostic Patterned Animal Re-identification by Aggregating Deep Local Features (IJCV 2024 · similarity 0.91)
- On Combining Animal Re-Identification Models to Address Small Datasets (IJCV 2026 · similarity 0.91)
- Toward Re-Identifying Any Animal (NeurIPS 2023 · similarity 0.90)
- RealBirdID: Benchmarking Bird Species Identification in the Era of MLLMs (CVPR 2026 · similarity 0.86)
- Generating the Wild: Individual-Consistent Image-to-Video Generation for Wildlife (NeurIPS 2026 · similarity 0.86)
- BuckTales: A multi-UAV dataset for multi-object tracking and re-identification of wild antelopes (NeurIPS 2024 · similarity 0.85)
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.
