AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

CoMatch: Dynamic Covisibility-Aware Transformer for Bilateral Subpixel-Level Semi-Dense Image Matching

Zizhuo Li, Yifan Lu, Linfeng Tang, Shihua Zhang, Jiayi Ma

ICCV 2025 · Architectures and efficiency › Attention and transformers · approach: Efficient image transformers · 6 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
add CGTC token condenser (vs depth-wise conv + max pooling) (ablation)CoMatch · ScanNetAUC@5deg: 20.0 -> 21.8Improved
use CAA instead of vanilla attention (ablation)CoMatch · ScanNetAUC@5deg: 20.7 -> 21.8Improved
use CAA instead of flash attention (ablation)CoMatch · ScanNetAUC@5deg: 20.5 -> 21.8Improved
use CAA instead of linear attention (ablation)CoMatch · ScanNetAUC@5deg: 18.1 -> 21.8Improved
use proposed 2nd-stage bilateral refinement instead of LoFTR's refinement (ablation)CoMatch · ScanNetAUC@5deg: 20.4 -> 21.8Improved
add 2nd-stage refinement (ablation)CoMatch · ScanNetAUC@5deg: 20.2 -> 21.8Improved
add the BSR module (ablation)CoMatch · ScanNetAUC@5deg: 11.3 -> 21.8Improved
use proposed backbone instead of RepVGG (ablation)CoMatch · ScanNetAUC@5deg: 21.2 -> 21.8No clear effect
use proposed backbone instead of ResNet-FPN (ablation)CoMatch · ScanNetAUC@5deg: 21.5 -> 21.8No clear effect
condensing range 2x2 instead of 4x4 (ablation)CoMatch · MegaDepth-1500runtime (ms): 123.8 -> 207.6Hurt
increase test resolution from 800x800 to 960x960CoMatch · MegaDepthAUC@5deg: 55.5 -> 56.3Improved
use CoMatch instead of ASpanFormersemi-dense matcher · MegaDepth-1500runtime (ms): 211.8 -> 123.8Improved

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