AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

CAIT: Triple-Win C ompression Toward High A ccuracy, Fast I nference, and Favorable T ransferability for ViTs

Ao Wang, Hui Chen, Zijia Lin, Sicheng Zhao, Jungong Han, Guiguang Ding

TPAMI 2025 · Architectures and efficiency › Attention and transformers · approach: Vision transformer pruning

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
add asymmetric merging (HTM and VTM alternating) instead of symmetric merging (ablation)ATME · ImageNet DeiT-TinyTop-1: 71.2 -> 71.9Improved
add head-level consistency (ablation)CDCP · ImageNet DeiT-TinyTop-1: 71.3 -> 72.7Improved
add attention-level consistency (ablation)CDCP · ImageNet DeiT-TinyTop-1: 72.1 -> 72.7Improved
add both consistencies (ablation)CDCP · ImageNet DeiT-TinyTop-1: 71.0 -> 72.7Improved
combine token merging with channel pruning (ablation)ATME · ImageNet DeiT-SmallTop-1: 79.8 -> 80.2Improved
fine-tune from pretrained instead of training ATME from scratch (ablation)ATME · ImageNet DeiT-TinyTop-1: 71.1 -> 71.9Improved
lengthen fine-tuning from 30 to 300 epochs (ablation)CAIT · ImageNet DeiT-SmallTop-1: 76.7 -> 80.2Improved
change warmup epochs before raising target ratio from 0 to 30 (ablation)CDCP · ImageNet DeiT-TinyTop-1: 72.26 -> 72.34No clear effect
change reconstruction interval iterations from 25 to 75 (ablation)CDCP · ImageNet DeiT-TinyTop-1: 72.34 -> 72.46No clear effect
use HTM only instead of alternating merging (ablation)ATME · ImageNet DeiT-TinyTop-1: 71.9 -> 71.5Hurt
use structured token merging instead of unstructured pruning for transferATME · ADE20k Semantic FPNmIoU: 42.7 -> 44.5Improved
use ATME instead of STViT-R reproducedATME · ImageNet Swin-TinyTop-1: 80.5 -> 81.1Improved

Later papers that compared with this method

1 later accepted paper reports CaiT as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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