{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/tracking-anything-with-decoupled-video","title":"Tracking Anything with Decoupled Video Segmentation","arxiv_id":"2309.03903","date":"2023-09-07","proceeding":"ICCV 2023 1","authors":["Ho Kei Cheng","Seoung Wug Oh","Brian Price","Alexander Schwing","Joon-Young Lee"],"abstract":"Training data for video segmentation are expensive to annotate. This impedes extensions of end-to-end algorithms to new video segmentation tasks, especially in large-vocabulary settings. To 'track anything' without training on video data for every individual task, we develop a decoupled video segmentation approach (DEVA), composed of task-specific image-level segmentation and class/task-agnostic bi-directional temporal propagation. Due to this design, we only need an image-level model for the target task (which is cheaper to train) and a universal temporal propagation model which is trained once and generalizes across tasks. To effectively combine these two modules, we use bi-directional propagation for (semi-)online fusion of segmentation hypotheses from different frames to generate a coherent segmentation. We show that this decoupled formulation compares favorably to end-to-end approaches in several data-scarce tasks including large-vocabulary video panoptic segmentation, open-world video segmentation, referring video segmentation, and unsupervised video object segmentation. Code is available at: https://hkchengrex.github.io/Tracking-Anything-with-DEVA","url_abs":"https://arxiv.org/abs/2309.03903v1","url_pdf":"https://arxiv.org/pdf/2309.03903v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"tracking-anything-with-decoupled-video","repo_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"open-vocabulary-video-segmentation","task_name":"Open-Vocabulary Video Segmentation"},{"task_slug":"open-world-video-segmentation","task_name":"Open-World Video Segmentation"},{"task_slug":"panoptic-segmentation","task_name":"Panoptic Segmentation"},{"task_slug":"referring-expression-segmentation","task_name":"Referring Expression Segmentation"},{"task_slug":"referring-video-object-segmentation","task_name":"Referring Video Object Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"semi-supervised-video-object-segmentation","task_name":"Semi-Supervised Video Object Segmentation"},{"task_slug":"unsupervised-video-object-segmentation","task_name":"Unsupervised Video Object Segmentation"},{"task_slug":"video-object-segmentation","task_name":"Video Object Segmentation"},{"task_slug":"video-panoptic-segmentation","task_name":"Video Panoptic Segmentation"},{"task_slug":"video-segmentation","task_name":"Video Segmentation"},{"task_slug":"video-semantic-segmentation","task_name":"Video Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/open-world-video-segmentation-on-burst-val","task":"Open-World Video Segmentation","dataset":"BURST-val","model":"DEVA (Mask2Former)","rank_in_archive_order":1,"of":3,"metrics":{"OWTA (all)":"69.9","OWTA (com)":"75.2","OWTA (unc)":"41.5"},"uses_additional_data":true},{"leaderboard":"/sota/open-world-video-segmentation-on-burst-val","task":"Open-World Video Segmentation","dataset":"BURST-val","model":"DEVA (EntitySeg)","rank_in_archive_order":2,"of":3,"metrics":{"OWTA (all)":"69.5","OWTA (com)":"73.3","OWTA (unc)":"50.5"},"uses_additional_data":true},{"leaderboard":"/sota/referring-expression-segmentation-on-davis","task":"Referring Expression Segmentation","dataset":"DAVIS 2017 (val)","model":"DEVA (ReferFormer)","rank_in_archive_order":3,"of":18,"metrics":{"J&F 1st frame":"66.3"},"uses_additional_data":true},{"leaderboard":"/sota/referring-expression-segmentation-on-refer-1","task":"Referring Expression Segmentation","dataset":"Refer-YouTube-VOS (2021 public validation)","model":"DEVA (ReferFormer)","rank_in_archive_order":14,"of":33,"metrics":{"J&F":"66.0"},"uses_additional_data":true},{"leaderboard":"/sota/semi-supervised-video-object-segmentation-on-1","task":"Semi-Supervised Video Object Segmentation","dataset":"DAVIS 2017 (test-dev)","model":"DEVA","rank_in_archive_order":8,"of":59,"metrics":{"F-measure (Mean)":"86.8","FPS":"25.3","J&F":"83.2","Jaccard (Mean)":"79.6"},"uses_additional_data":true},{"leaderboard":"/sota/visual-object-tracking-on-davis-2017","task":"Semi-Supervised Video Object Segmentation","dataset":"DAVIS 2017 (val)","model":"DEVA","rank_in_archive_order":12,"of":81,"metrics":{"F-measure (Mean)":"91.0","J&F":"87.6","Jaccard (Mean)":"84.2","Speed (FPS)":"25.3"},"uses_additional_data":true},{"leaderboard":"/sota/semi-supervised-video-object-segmentation-on-21","task":"Semi-Supervised Video Object Segmentation","dataset":"MOSE","model":"DEVA (with OVIS)","rank_in_archive_order":8,"of":17,"metrics":{"F":"70.8","FPS":"25.3","J":"62.3","J&F":"66.5"},"uses_additional_data":false},{"leaderboard":"/sota/semi-supervised-video-object-segmentation-on-21","task":"Semi-Supervised Video Object Segmentation","dataset":"MOSE","model":"DEVA (no OVIS)","rank_in_archive_order":11,"of":17,"metrics":{"F":"64.3","FPS":"25.3","J":"55.8","J&F":"60.0"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-video-object-segmentation-on-10","task":"Unsupervised Video Object Segmentation","dataset":"DAVIS 2016 val","model":"DEVA (DIS)","rank_in_archive_order":2,"of":25,"metrics":{"F":"90.2","G":"88.9","J":"87.6"},"uses_additional_data":true},{"leaderboard":"/sota/unsupervised-video-object-segmentation-on-5","task":"Unsupervised Video Object Segmentation","dataset":"DAVIS 2017 (test-dev)","model":"DEVA (EntitySeg)","rank_in_archive_order":1,"of":6,"metrics":{"J&F":"62.1"},"uses_additional_data":true},{"leaderboard":"/sota/unsupervised-video-object-segmentation-on-4","task":"Unsupervised Video Object Segmentation","dataset":"DAVIS 2017 (val)","model":"DEVA (EntitySeg)","rank_in_archive_order":1,"of":10,"metrics":{"F-measure (Mean)":"76.4","J&F":"73.4","Jaccard (Mean)":"70.4"},"uses_additional_data":true},{"leaderboard":"/sota/video-panoptic-segmentation-on-vipseg","task":"Video Panoptic Segmentation","dataset":"VIPSeg","model":"DEVA (Mask2Former - SwinB)","rank_in_archive_order":6,"of":12,"metrics":{"STQ":"52.2","VPQ":"55.0"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2309.03903","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2309.03903"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"ran":7,"unverified":3},"by_repo_kind":{"official":{"samples":10,"ran":7,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":10,"samples":[{"code_sha256_prefix":"48f5a5ec1d5dd2ef","entry":"conv3x3","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/resnet.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/resnet.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"48f5a5ec1d5dd2ef"}},{"code_sha256_prefix":"0d2e632209183345","entry":"dice_loss","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/losses.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/losses.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"0d2e632209183345"}},{"code_sha256_prefix":"b8e2bcf17a7d1985","entry":"do_softmax","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/memory_utils.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/memory_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"b8e2bcf17a7d1985"}},{"code_sha256_prefix":"2048782602814ecb","entry":"downsample_groups","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/group_modules.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/group_modules.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2048782602814ecb"}},{"code_sha256_prefix":"14cdff9c8a6cd96f","entry":"get_similarity","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/memory_utils.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/memory_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"14cdff9c8a6cd96f"}},{"code_sha256_prefix":"b4da85ddeff2a196","entry":"resnet18","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/resnet.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/resnet.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"b4da85ddeff2a196"}},{"code_sha256_prefix":"0d0b892609b02b52","entry":"resnet50","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/resnet.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/resnet.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"0d0b892609b02b52"}},{"code_sha256_prefix":"8185b445f454d15e","entry":"get_affinity","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/memory_utils.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/memory_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"8185b445f454d15e"}},{"code_sha256_prefix":"abf061dab6b3f9b9","entry":"interpolate_groups","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/group_modules.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/group_modules.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"abf061dab6b3f9b9"}},{"code_sha256_prefix":"89405f8a57f3c3a5","entry":"upsample_groups","repo":"hkchengrex/Tracking-Anything-with-DEVA","repo_kind":"official","path":"deva/model/group_modules.py","file_url":"https://github.com/hkchengrex/Tracking-Anything-with-DEVA/blob/HEAD/deva/model/group_modules.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"89405f8a57f3c3a5"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}