Research map / Label-efficient and robust learning / Recognition and long-tail
Free-Grained Hierarchical Visual Recognition
Seulki Park, Zilin Wang, Stella X. Yu
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 |
|---|---|---|---|
| train hierarchical classifier with free-grain labels instead of full labels | H-CAST · CUB-F | Full-path accuracy (%): 84.9 -> 45.1 | Hurt |
| train hierarchical classifier with free-grain labels instead of full labels | H-CAST · iNat21-mini-F | Full-path accuracy (%): 64.9 -> 25.6 | Hurt |
| add text-guided pseudo attributes (contrastive image-text loss) to H-CAST | H-CAST · ImageNet-F | Full-path accuracy (%): 57.59 -> 63.20 | Improved |
| add text-guided pseudo attributes to H-CAST | H-CAST · iNat21-mini-F | Full-path accuracy (%): 25.63 -> 29.74 | Improved |
| add taxonomy-guided semi-supervised learning (Taxon-SSL) instead of H-CAST | H-ViT · iNat21-mini-F | Full-path accuracy (%): 25.63 -> 31.74 | Improved |
| use Taxon-SSL instead of H-CAST | H-ViT · ImageNet-F | Full-path accuracy (%): 57.59 -> 48.40 | Hurt |
| combine Taxon-SSL with Text-Attr (ablation) | Taxon-SSL · ImageNet-F | Full-path accuracy (%): 48.40 -> 49.65 | Improved |
| combine Taxon-SSL with Text-Attr (ablation) | Taxon-SSL · iNat21-mini-F | Full-path accuracy (%): 31.74 -> 31.93 | Improved |
| use Text-Attr rather than Taxon-SSL when fine-grained labels are scarce | ImageNet-F classes · ImageNet-F per-class fine accuracy | per-class accuracy | Mixed |
| use consistency-based stopping instead of confidence-based stopping (tau=0.9) (ablation) | Free-grained inference, Text-Attr (H-CAST) · ImageNet-F | Fine-level reached and correct (%): 48% -> 63% | Improved |
| add text-guided pseudo attributes to H-CAST on ImageNet-F | H-CAST · ImageNet-F | TICE (%): 21.81 -> 18.58 | Improved |
Closest papers by meaning
- Democratizing Fine-grained Visual Recognition with Large Language Models (ICLR 2024 · similarity 0.89)
- Learning Hierarchical Image Segmentation For Recognition and By Recognition (ICLR 2024 · similarity 0.88)
- From Coarse to Fine-Grained Open-Set Recognition (CVPR 2024 · similarity 0.88)
- Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: A Comprehensive Evaluation (ICLR 2026 · similarity 0.86)
- Fine-grained Classes and How to Find Them (ICML 2024 · similarity 0.86)
- Fine-R1: Make Multi-modal LLMs Excel in Fine-Grained Visual Recognition by Chain-of-Thought Reasoning (ICLR 2026 · 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.
