Research map / Label-efficient and robust learning / Recognition and long-tail
Learning to Navigate for Fine-grained Classification
Ze Yang, Tiange Luo, Dong Wang, Zhiqiang Hu, Jun Gao, Liwei Wang
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 Teacher guidance to Navigator (ablation) | NTS-Net (K=4) · CUB-200-2011 | top-1 accuracy: 83.3% -> 87.5% | Improved |
| add informative regions (K=2) to full image input (ablation) | NTS-Net · CUB-200-2011 | top-1 accuracy: 85.3% -> 87.3% | Improved |
| increase number of regions K from 2 to 4 (ablation) | NTS-Net · CUB-200-2011 | top-1 accuracy: 87.3% -> 87.5% | No clear effect |
| increase number of regions K from 0 to 4 (ablation) | NTS-Net · CUB-200-2011 | top-1 accuracy: 85.3% -> 87.5% | Improved |
| use NTS-Net full-image-only (K=0) instead of plain ResNet-50 (ablation) | ResNet-50 · CUB-200-2011 | top-1 accuracy: 84.5% -> 85.3% | Improved |
| use NTS-Net (K=4) instead of ResNet-50 | ResNet-50 · CUB-200-2011 | top-1 accuracy: 84.5% -> 87.5% | Improved |
| increase K from 2 to 4 (ablation) | NTS-Net · FGVC Aircraft | top-1 accuracy: 90.8% -> 91.4% | Improved |
| increase K from 2 to 4 (ablation) | NTS-Net · Stanford Cars | top-1 accuracy: 93.7% -> 93.9% | Improved |
| use NTS-Net instead of MA-CNN | MA-CNN · FGVC Aircraft | top-1 accuracy: 89.9% -> 91.4% | Improved |
| use NTS-Net instead of DT-RAM | DT-RAM · Stanford Cars | top-1 accuracy: 93.1% -> 93.9% | Improved |
Closest papers by meaning
- Seeing as Experts Do: A Knowledge-Augmented Agent for Open-Set Fine-Grained Visual Understanding (CVPR 2026 · similarity 0.85)
- GranViT: A Fine-Grained Vision Model With Autoregressive Perception For MLLMs (ICLR 2026 · similarity 0.84)
- Democratizing Fine-grained Visual Recognition with Large Language Models (ICLR 2024 · similarity 0.84)
- Fine-grained Classes and How to Find Them (ICML 2024 · similarity 0.84)
- Specificity-aware reinforcement learning for fine-grained open-world classification (CVPR 2026 · similarity 0.83)
- Learning State-Based Node Representations from a Class Hierarchy for Fine-Grained Open-Set Detection (ICML 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.
