Occluded video instance segmentation: A benchmark
Can our video understanding systems perceive objects when a heavy occlusion exists in a scene? To answer this question, we collect a large-scale dataset called OVIS for occluded video instance segmentation, that is, to simultaneously detect, segment, and track instances in occluded scenes. OVIS cons...
Main Authors: | , , , , , , , , , |
---|---|
Format: | Journal article |
Language: | English |
Published: |
Springer
2022
|
_version_ | 1797107392973897728 |
---|---|
author | Qi, J Gao, Y Hu, Y Wang, X Liu, X Bai, X Belongie, S Yuille, A Torr, PHS Bai, S |
author_facet | Qi, J Gao, Y Hu, Y Wang, X Liu, X Bai, X Belongie, S Yuille, A Torr, PHS Bai, S |
author_sort | Qi, J |
collection | OXFORD |
description | Can our video understanding systems perceive objects when a heavy occlusion exists in a scene? To answer this question, we collect a large-scale dataset called OVIS for occluded video instance segmentation, that is, to simultaneously detect, segment, and track instances in occluded scenes. OVIS consists of 296k high-quality instance masks from 25 semantic categories, where object occlusions usually occur. While our human vision systems can understand those occluded instances by contextual reasoning and association, our experiments suggest that current video understanding systems cannot. On the OVIS dataset, the highest AP achieved by state-of-the-art algorithms is only 16.3, which reveals that we are still at a nascent stage for understanding objects, instances, and videos in a real-world scenario. We also present a simple plug-and-play module that performs temporal feature calibration to complement missing object cues caused by occlusion. Built upon MaskTrack R-CNN and SipMask, we obtain a remarkable AP improvement on the OVIS dataset. The OVIS dataset and project code are available at http://songbai.site/ovis. |
first_indexed | 2024-03-07T07:15:23Z |
format | Journal article |
id | oxford-uuid:775c0679-5072-4f19-bd82-2c2e95a55a30 |
institution | University of Oxford |
language | English |
last_indexed | 2024-03-07T07:15:23Z |
publishDate | 2022 |
publisher | Springer |
record_format | dspace |
spelling | oxford-uuid:775c0679-5072-4f19-bd82-2c2e95a55a302022-08-04T06:34:55ZOccluded video instance segmentation: A benchmarkJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:775c0679-5072-4f19-bd82-2c2e95a55a30EnglishSymplectic ElementsSpringer2022Qi, JGao, YHu, YWang, XLiu, XBai, XBelongie, SYuille, ATorr, PHSBai, SCan our video understanding systems perceive objects when a heavy occlusion exists in a scene? To answer this question, we collect a large-scale dataset called OVIS for occluded video instance segmentation, that is, to simultaneously detect, segment, and track instances in occluded scenes. OVIS consists of 296k high-quality instance masks from 25 semantic categories, where object occlusions usually occur. While our human vision systems can understand those occluded instances by contextual reasoning and association, our experiments suggest that current video understanding systems cannot. On the OVIS dataset, the highest AP achieved by state-of-the-art algorithms is only 16.3, which reveals that we are still at a nascent stage for understanding objects, instances, and videos in a real-world scenario. We also present a simple plug-and-play module that performs temporal feature calibration to complement missing object cues caused by occlusion. Built upon MaskTrack R-CNN and SipMask, we obtain a remarkable AP improvement on the OVIS dataset. The OVIS dataset and project code are available at http://songbai.site/ovis. |
spellingShingle | Qi, J Gao, Y Hu, Y Wang, X Liu, X Bai, X Belongie, S Yuille, A Torr, PHS Bai, S Occluded video instance segmentation: A benchmark |
title | Occluded video instance segmentation: A benchmark |
title_full | Occluded video instance segmentation: A benchmark |
title_fullStr | Occluded video instance segmentation: A benchmark |
title_full_unstemmed | Occluded video instance segmentation: A benchmark |
title_short | Occluded video instance segmentation: A benchmark |
title_sort | occluded video instance segmentation a benchmark |
work_keys_str_mv | AT qij occludedvideoinstancesegmentationabenchmark AT gaoy occludedvideoinstancesegmentationabenchmark AT huy occludedvideoinstancesegmentationabenchmark AT wangx occludedvideoinstancesegmentationabenchmark AT liux occludedvideoinstancesegmentationabenchmark AT baix occludedvideoinstancesegmentationabenchmark AT belongies occludedvideoinstancesegmentationabenchmark AT yuillea occludedvideoinstancesegmentationabenchmark AT torrphs occludedvideoinstancesegmentationabenchmark AT bais occludedvideoinstancesegmentationabenchmark |