Papers › Separable Structure Modeling for Semi-supervised Video Object Segmentation

Separable Structure Modeling for Semi-supervised Video Object Segmentation

18 Feb 2021archive 2025-07-28

Wencheng Zhu, Jiahao Li, Jiwen Lu, Jie zhou

In this paper, we propose a separable structure modeling approach for semi-supervised video object segmentation. Unlike most existing methods which preclude the semantically structural information of target objects, our method not only captures pixel-level similarity relationships between the reference and target frames but also reveals the separable structure of the specified objects in target frames. Specifically, we first compute a pixel-wise similarity matrix by using representations of reference and target pixels and then select top-rank reference pixels for target pixel classification. According to the prior knowledge from these top-rank reference pixels, we further appoint the representative target pixels for object structure modeling. Particularly, in the structure modeling branch, we extract the shared and individual features that can well represent the whole object and its components, respectively. Moreover, the proposed method is a fast algorithm without online fine-tuning and any post-processing. We conduct extensive experiments and ablation studies on the DAVIS-16, DAVIS-17, and YouTube-VOS datasets, and experimental results on three widely-used datasets demonstrate that our method achieves superior performance, compared with state-of-the-art semi-supervised video object segmentation approaches in terms of speed and accuracy.

PaperPDFCode

Code

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

ObjectOne-shot visual object segmentationSemi-Supervised Video Object SegmentationVideo Object SegmentationVideo Semantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS F-measure (Decay) 5.6 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS F-measure (Mean) 85.6 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS F-measure (Recall) 92.3 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS J&F 85.9 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS Jaccard (Decay) 5.3 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS Jaccard (Mean) 86.2 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS Jaccard (Recall) 97.1 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2016 SSM-VOS Speed (FPS) 36.5 #49 of 78 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) SSM-VOS F-measure (Decay) 25.3 #46 of 59 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) SSM-VOS F-measure (Mean) 63.8 #46 of 59 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) SSM-VOS J&F 62.0 #46 of 59 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) SSM-VOS Jaccard (Decay) 23.5 #46 of 59 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (test-dev) SSM-VOS Jaccard (Mean) 60.2 #46 of 59 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS F-measure (Decay) 15.3 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS F-measure (Mean) 79.9 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS J&F 77.6 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS Jaccard (Decay) 11.7 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS Jaccard (Mean) 75.3 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation DAVIS 2017 (val) SSM-VOS Speed (FPS) 22.3 #49 of 81 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS F-Measure (Seen) 73.3 #47 of 53 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS F-Measure (Unseen) 62.6 #47 of 53 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS Jaccard (Seen) 72.3 #47 of 53 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS Jaccard (Unseen) 57.8 #47 of 53 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS Overall 66.5 #47 of 53 Archive leaderboard report
Semi-Supervised Video Object Segmentation YouTube-VOS 2018 SSM-VOS Speed (FPS) 24.1 #47 of 53 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.

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