AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

CF-ViT: A General Coarse-to-Fine Method for Vision Transformer

Mengzhao Chen, Mingbao Lin, Ke Li, Yunhang Shen, Yongjian Wu, Fei Chao, Rongrong Ji

AAAI 2023 · Architectures and efficiency › Attention and transformers · approach: Hybrid and interpretable ViT designs · 70 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use coarse-to-fine inference with η=0.5DeiT-S · ImageNet valFLOPs (G): 4.6 -> 1.8Improved
Use coarse-to-fine inference with η=0.63LV-ViT-S · ImageNet valFLOPs (G): 6.6 -> 3.1Improved
Use coarse-to-fine inference with η=0.5DeiT-S · ImageNet valTop-1 accuracy (%): 79.8 -> 79.8No clear effect
Use global class attention instead of last-layer attention for region selection (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.3 -> 80.8Improved
Use informative global class-attention regions rather than random regions (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 79.6 -> 80.8Improved
Select high-attention rather than low-attention regions for fine splitting (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 77.6 -> 80.8Improved
Add feature reuse to the fine-stage tokens (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.0 -> 80.8Improved
Exclude the [class] token from feature reuse (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.2 -> 80.8Improved
Exclude uninformative tokens from feature reuse (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.6 -> 80.8Improved
Use an MLP rather than a linear layer for feature reuse (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.6 -> 80.8Improved
Increase the informative-patch rate α (ablation)CF-ViT · ImageNet, fine inferenceTop-1 accuracy (%): 80.4 -> 81.4Improved
Increase the informative-patch rate α (ablation)CF-ViT · ImageNet, fine inferenceFLOPs (G): 3.7 -> 5.5Hurt

Later papers that compared with this method

2 later accepted papers report CF-ViT as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy