AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Frequency-Aware Affinity for Weakly Supervised Semantic Segmentation

Ziqian Yang, Xianglin Qiu, Xinqiao Zhao, Xiaolei Wang, Quan Zhang, Jimin Xiao

CVPR 2026 · Detection and segmentation › Weakly supervised detection and segmentation · approach: CAM refinement and prototypes

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 inverse (high-pass) affinity to traditional CAM refinement (ablation)smoothed affinity refinement · PASCAL VOC 2012 trainCAM mIoU: 64.8 -> 67.2Improved
add alignment and rectification losses to produce Ala and Aha (ablation)traditional refinement · PASCAL VOC 2012 trainCAM mIoU: 67.2 -> 77.9Improved
replace traditional element-wise refinement with Frequency-Guided OT CAM generation (ablation)DFA · PASCAL VOC 2012 trainCAM mIoU: 77.9 -> 79.1Improved
use high-pass affinity alone for CAM refinement (ablation)CAM refinement · PASCAL VOC 2012 trainqualitative CAM qualityMixed
use low-pass (self-attention) affinity alone for CAM refinement (ablation)CAM refinement · PASCAL VOC 2012 trainqualitative CAM qualityMixed
DFA vs ExCEL on CAM seedsDFA · PASCAL VOC 2012 trainCAM mIoU: 78.0 -> 79.1Improved
DFA vs ExCEL on final segmentationDFA · MS COCO 2014 valmIoU: 50.3 -> 51.5Improved

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