Research map / Detection and segmentation / CNN features
Fine-Grained Species Recognition With Privileged Pooling: Better Sample Efficiency Through Supervised Attention
Andrés C. Rodríguez, Stefano D’Aronco, Konrad Schindler B, Jan Dirk Wegner
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 keypoint supervision to attention maps (ablation) | AvgPool PrPool · CUB200 | top-1 accuracy: 84.3 -> 88.1 | Improved |
| add keypoint supervision to attention maps (ablation) | CovPool PrPool · CUB200 | top-1 accuracy: 87.1 -> 89.0 | Improved |
| add keypoint supervision to attention maps (ablation) | AvgPool PrPool · iBirds2017 | top-1 accuracy: 60.3 -> 65.2 | Improved |
| add keypoint supervision to attention maps (ablation) | CovPool PrPool · iBirds2017 | top-1 accuracy: 62.3 -> 67.0 | Improved |
| add complementary-map regulariser alone (no keypoint supervision) (ablation) | AvgPool attention · CUB200 | top-1 accuracy: 84.3 -> 87.0 | Improved |
| add attention-cropped re-feeding at test time (ablation) | AvgPrPool · CUB200 | top-1 accuracy: 87.7 -> 88.1 | Improved |
| add attention-cropped re-feeding at test time (ablation) | WSDAN · iBirds2017 | top-1 accuracy: 54.9 -> 62.4 | Improved |
| add attention-cropped re-feeding at test time (ablation) | AvgPrPool · CCT20+ Cis locations | overall accuracy: 81.0 -> 80.3 | Hurt |
| add attention-cropped re-feeding at test time (ablation) | CovPrPool · CCT20+ Cis locations | overall accuracy: 76.7 -> 68.8 | Hurt |
| increase attention maps from 15 to 32 (add 17 complementary) (ablation) | CovPrPool · CUB200 | top-1 accuracy: 85.6 -> 89.2 | Improved |
| increase attention maps from 32 to 64 (ablation) | CovPrPool · CUB200 | top-1 accuracy: 89.2 -> 88.9 | Hurt |
| use covariance pooling instead of average pooling with privileged attention | PrPool · CUB200 | top-1 accuracy: 87.7 -> 89.2 | Improved |
Closest papers by meaning
- Prompt-CAM: Making Vision Transformers Interpretable for Fine-Grained Analysis (CVPR 2025 · similarity 0.85)
- Fish-Vista: A Multi-Purpose Dataset for Understanding & Identification of Traits from Images (CVPR 2025 · similarity 0.84)
- BioTrove: A Large Curated Image Dataset Enabling AI for Biodiversity (NeurIPS 2024 · similarity 0.84)
- Multi-scale Activation, Refinement, and Aggregation: Exploring Diverse Cues for Fine-Grained Bird Recognition (AAAI 2025 · similarity 0.83)
- UniAP: Towards Universal Animal Perception in Vision via Few-Shot Learning (AAAI 2024 · similarity 0.83)
- Open-Insect: Benchmarking Open-Set Recognition of Novel Species in Biodiversity Monitoring (NeurIPS 2025 · similarity 0.83)
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.
