AI Paper Atlas

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

Learning to Detect Instance-Level Salient Objects Using Complementary Image Labels

Xin Tian, Ke Xu, Xin She Yang, Baocai Yin, Rynson W. H. Lau

IJCV 2022 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Weakly supervised detection · 27 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 Cross-layer Attention module (ablation)WSID-Net · ILSO testmAP@0.5: 64.3% -> 68.3%Improved
add Boundary Enhancement module (ablation)WSID-Net · ILSO testmAP@0.5: 65.2% -> 68.3%Improved
add centroid-based subitizing loss (ablation)WSID-Net · ILSO testmAP@0.7: 46.9% -> 51.7%Improved
add Progressive Training Scheme (ablation)WSID-Net · ILSO testmAP@0.7: 47.2% -> 51.7%Improved
add all four components (CA, BE, PTS, L_SU) (ablation)WSID-Net · ILSO testmAP@0.7: 37.4% -> 51.7%Improved
add cross-layer feature mixing to CA (ablation)CA module · ILSO testmAP@0.5: 66.1% -> 68.3%Improved
add parallel mutual attention units to CA (ablation)CA module · ILSO testmAP@0.5: 66.8% -> 68.3%Improved
parallel instead of cascade (s->c) spatial/channel attention (ablation)MA unit · ILSO testmAP@0.5: 67.4% -> 68.3%Improved
add Canny filter to BE module (ablation)BE module · ILSO testmAP@0.5: 65.0% -> 68.3%Improved
combine learned boundary features with Canny instead of Canny only (ablation)BE module · ILSO testmAP@0.7: 48.0% -> 51.7%Improved
automatic median-based Canny thresholds instead of manual 30/200 (ablation)BE module · ILSO testmAP@0.7: 50.8% -> 51.7%Improved
automatic median-based Canny thresholds instead of manual 230/260 (ablation)BE module · ILSO testmAP@0.7: 50.2% -> 51.7%Improved

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