AI Paper Atlas

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

Concealed Object Detection

Deng-Ping Fan, Ge-Peng Ji, Ming‐Ming Cheng, Ling Shao

TPAMI 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Open-set and unknown object detection · 842 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 neighbor connection decoder (NCD) (ablation)SINet · CAMO-TestE-measure: 0.869 -> 0.882Improved
use NCD instead of partial decoder (ablation)SINet · CHAMELEONweighted F-measure: 0.799 -> 0.816Improved
add texture enhanced module (TEM) (ablation)SINet · CAMO-TestE-measure: 0.867 -> 0.882Improved
use asymmetric instead of symmetric convolutions in TEM (ablation)SINet · CAMO-TestE-measure: 0.866 -> 0.882Improved
reverse guidance only in first GRA block instead of none (ablation)SINet · CAMO-Testweighted F-measure: 0.739 -> 0.743Improved
reverse guidance in first two blocks instead of only first (ablation)SINet · COD10K-Testweighted F-measure: 0.680 -> 0.661Hurt
reverse guidance in all three blocks instead of only first (ablation)SINet · COD10K-Testweighted F-measure: 0.680 -> 0.670Hurt
progressive group size {32;8;1} instead of uniform {32;32;32} (ablation)SINet · COD10K-Testweighted F-measure: 0.675 -> 0.680Improved
progressive group size {32;8;1} instead of uniform {1;1;1} (ablation)SINet · COD10K-Testweighted F-measure: 0.681 -> 0.680No clear effect
use COD10K instead of CAMO as training set for cross-dataset generalizationSINet · COD10K/CAMOS-measure drop: 12.6% -> -6.0%Improved
use SINet instead of SINet_cvprSINet · COD10K-Testweighted F-measure: 0.551 -> 0.680Improved

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