Papers › Group Collaborative Learning for Co-Salient Object Detection

Group Collaborative Learning for Co-Salient Object Detection

15 Mar 2021CVPR 2021 1arXiv:2104.01108archive 2025-07-28

Qi Fan, Deng-Ping Fan, Huazhu Fu, Chi Keung Tang, Ling Shao, Yu-Wing Tai

We present a novel group collaborative learning framework (GCoNet) capable of detecting co-salient objects in real time (16ms), by simultaneously mining consensus representations at group level based on the two necessary criteria: 1) intra-group compactness to better formulate the consistency among co-salient objects by capturing their inherent shared attributes using our novel group affinity module; 2) inter-group separability to effectively suppress the influence of noisy objects on the output by introducing our new group collaborating module conditioning the inconsistent consensus. To learn a better embedding space without extra computational overhead, we explicitly employ auxiliary classification supervision. Extensive experiments on three challenging benchmarks, i.e., CoCA, CoSOD3k, and Cosal2015, demonstrate that our simple GCoNet outperforms 10 cutting-edge models and achieves the new state-of-the-art. We demonstrate this paper's new technical contributions on a number of important downstream computer vision applications including content aware co-segmentation, co-localization based automatic thumbnails, etc.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

fanq15/GCoNet officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Co-Salient Object DetectionObjectObject DetectionSalient Object Detectionobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Co-Salient Object Detection CoCA GCoNet MAE 0.105 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoCA GCoNet Mean F-measure 0.531 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoCA GCoNet S-measure 0.673 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoCA GCoNet max E-measure 0.760 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoCA GCoNet max F-measure 0.544 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoCA GCoNet mean E-measure 0.739 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet MAE 0.071 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet S-measure 0.802 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet max E-measure 0.860 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet max F-measure 0.777 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet mean E-measure 0.857 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSOD3k GCoNet mean F-measure 0.770 #5 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet MAE 0.068 #6 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet S-measure 0.845 #6 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet max E-measure 0.888 #6 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet max F-measure 0.847 #6 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet mean E-measure 0.884 #6 of 10 Archive leaderboard report
Co-Salient Object Detection CoSal2015 GCoNet mean F-measure 0.838 #6 of 10 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AWARE

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections