AI Paper Atlas

Research map / Detection and segmentation / CNN features

Mutual Graph Learning for Camouflaged Object Detection

Qiang Zhai, Xin Li, Fan Yang, Chenglizhao Chen, Hong Cheng, Deng-Ping Fan

CVPR 2021 · Detection and segmentation › CNN features · approach: Context for hard detection conditions · 349 citations (OpenAlex)

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 mutual graph learning with COD and edge extractionCamouflaged object detection · CHAMELEONS-measure: 0.869 -> 0.893Improved
Use mutual graph learning with COD and edge extractionCamouflaged object detection · CAMO-TestS-measure: 0.769 -> 0.775Improved
Use mutual graph learning with COD and edge extractionCamouflaged object detection · COD10K-TestS-measure: 0.789 -> 0.814Improved
Add RIGR to the ResNet-50 FCN baseline (ablation)MGL · CAMO-TestS-measure: 0.742 -> 0.766Improved
Add ECGR to ResNet-50 FCN + RIGR (ablation)MGL · CAMO-TestS-measure: 0.766 -> 0.772Improved
Use K=16 semantic nodes instead of K=32 (ablation)S-MGL · CAMO-TestS-measure: 0.771 -> 0.772No clear effect
Use K=64 semantic nodes instead of K=32 (ablation)S-MGL · CAMO-TestS-measure: 0.772 -> 0.774No clear effect
Use z=16 edge supportive nodes instead of z=32 (ablation)S-MGL · CAMO-TestS-measure: 0.772 -> 0.772No clear effect
Use z=64 edge supportive nodes instead of z=32 (ablation)S-MGL · CAMO-TestS-measure: 0.772 -> 0.773No clear effect
Use two recurrent stages instead of one (ablation)R-MGL · CAMO-TestS-measure: 0.772 -> 0.775Improved
Use three recurrent stages instead of two (ablation)R-MGL · CAMO-TestS-measure: 0.775 -> 0.773Hurt

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