AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

CoGenSAM: Codebook-Interactive Generative Labeling for Adapting SAM to Crack Segmentation

Zhuangzhuang Chen, Nuo Chen, Dachong Li, Zhiliang Lin, Xingyu Feng, Yifan Zhang, Jie Chen, Jianqiang Li

AAAI 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Interactive and referring segmentation · 1 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 the restoration model to edge-detection labels (ablation)CoGenSAM · DeepCrackmIoU: 49.22 -> 61.73Improved
add the restoration model to edge-detection labels (ablation)CoGenSAM · Bridge2025mIoU: 32.78 -> 48.25Improved
add the pre-trained crack-aware codebook (ablation)CoGenSAM · DeepCrackmIoU: 57.33 -> 61.73Improved
add the pre-trained crack-aware codebook (ablation)CoGenSAM · Bridge2025mIoU: 43.98 -> 48.25Improved
add contrastive-interaction loss Lc (ablation)CoGenSAM · DeepCrackmIoU: 58.01 -> 61.73Improved
add contrastive-interaction loss Lc (ablation)CoGenSAM · Bridge2025mIoU: 44.52 -> 48.25Improved
add contrastive-interaction loss Lc (ablation)CoGenSAM · DeepCrackF1: 71.41 -> 74.35Improved
swap Sobel for Roberts, Prewitt or Canny edge detection (ablation)CoGenSAM · DeepCrackmIoU: 61.57 -> 61.34 / 61.00 / 60.41No clear effect
use generated labels instead of manual annotation (annotation-free vs CrackSAM)SAM · DeepCrackmIoU: 65.62 -> 61.57Hurt
use generated labels instead of manual annotation (annotation-free vs CrackSAM)SAM · Crack500mIoU: 54.22 -> 46.33Hurt
use CoGenSAM generated labels instead of FreeCOS annotation-free methodSAM · Bridge2025mIoU: 41.05 -> 48.12Improved
use CoGenSAM instead of Up-CrackNetSAM · Crack500mIoU: 21.27 -> 46.33Improved

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