Research map / Label-efficient and robust learning / Recognition and long-tail
Maximum-Entropy Fine Grained Classification
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 |
|---|---|---|---|
| maximize output prediction entropy during fine-tuning | CNN fine-tuning · CUB-200-2011 | Top-1: 68.19 -> 74.37 | Helped |
| maximize output prediction entropy during fine-tuning | CNN fine-tuning · Cars | Top-1: 91.52 -> 93.85 | Helped |
| maximize output prediction entropy during fine-tuning | CNN fine-tuning · Aircrafts | Top-1: 86.30 -> 89.76 | Helped |
| maximize output prediction entropy during fine-tuning | CNN fine-tuning · NABirds | Top-1: 80.90 -> 82.66 | Helped |
| maximize output prediction entropy during fine-tuning | CNN fine-tuning · Stanford Dogs | Top-1: 82.13 -> 83.18 | Helped |
| maximize output prediction entropy (ablation) | GoogLeNet · CIFAR-10 | accuracy: 84.16 -> 84.10 | Hurt |
| maximize output prediction entropy (ablation) | GoogLeNet · CIFAR-100 | accuracy: 70.24 -> 73.50 | Helped |
| maximize output prediction entropy (ablation) | DenseNet-121 · CIFAR-10 | accuracy: 92.19 -> 92.22 | No clear effect |
| maximize output prediction entropy (ablation) | DenseNet-121 · CIFAR-100 | accuracy: 75.01 -> 76.22 | Helped |
| apply maximum-entropy training to a lower-diversity dog-class subset rather than a random subset (ablation) | ImageNet subset classification · Dogs-ImageNet versus Random-ImageNet | accuracy gain over baseline: (0.35) -> (2.63); (0.28) -> (1.86) | Helped |
| use maximum-entropy training instead of label smoothing regularization (ablation) | VGGNet16 fine-tuning · CUB-200-2011 | Top-1: 70.03 -> 77.02 | Helped |
| use maximum-entropy training instead of label smoothing regularization (ablation) | VGGNet16 fine-tuning · Cars | Top-1: 81.45 -> 83.88 | Helped |
Closest papers by meaning
- Advancing Fine-Grained Classification by Structure and Subject Preserving Augmentation (NeurIPS 2024 · similarity 0.87)
- Novel Class Discovery for Ultra-Fine-Grained Visual Categorization (CVPR 2024 · similarity 0.86)
- Cross-Layer and Cross-Sample Feature Optimization Network for Few-Shot Fine-Grained Image Classification (AAAI 2024 · similarity 0.85)
- Twofold Debiasing Enhances Fine-Grained Learning with Coarse Labels (AAAI 2025 · similarity 0.85)
- 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)
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.