{"url":"/task/motion-segmentation","name":"Motion Segmentation","slug":"motion-segmentation","description_markdown":"**Motion Segmentation** is an essential task in many applications in Computer Vision and Robotics, such as surveillance, action recognition and scene understanding. The classic way to state the problem is the following: given a set of feature points that are tracked through a sequence of images, the goal is to cluster those trajectories according to the different motions they belong to. It is assumed that the scene contains multiple objects that are moving rigidly and independently in 3D-space.\n\n\n<span class=\"description-source\">Source: [Robust Motion Segmentation from Pairwise Matches ](https://arxiv.org/abs/1905.09043)</span>","categories":[{"name":"Computer Vision","url":"/area/computer-vision"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"archive_url"},"counts":{"papers_tagged":212,"papers_with_code":62,"benchmarks":4,"benchmark_tables_in_archive":4,"benchmark_tables_shown":4,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":7,"subtasks":0,"parent_tasks":0},"benchmarks":[{"leaderboard":"/sota/motion-segmentation-on-apolloscape","slug":"motion-segmentation-on-apolloscape","dataset":"ApolloScape","dataset_url":"/dataset/apolloscape-1","rows_in_archive":5,"metrics":["Accuracy"],"first_row_in_archive_order":{"model":"Rule Based","paper_title":"Understanding Dynamic Scenes using Graph Convolution Networks","paper_url":"/paper/understanding-dynamic-scenes-using-graph","paper_date":"2020-05-09","arxiv_id":"2005.04437","code_links":[{"title":"ma8sa/Undersrtanding-Dynamic-Scenes-using-MR-GCN","url":"https://github.com/ma8sa/Undersrtanding-Dynamic-Scenes-using-MR-GCN"}],"syntology":null}},{"leaderboard":"/sota/motion-segmentation-on-hopkins155","slug":"motion-segmentation-on-hopkins155","dataset":"Hopkins155","dataset_url":"/dataset/hopkins155","rows_in_archive":4,"metrics":["Classification Error"],"first_row_in_archive_order":{"model":"MVC","paper_title":"Motion Segmentation by Exploiting Complementary Geometric Models","paper_url":"/paper/motion-segmentation-by-exploiting","paper_date":"2018-04-06","arxiv_id":"1804.02142","code_links":[{"title":"alex-xun-xu/MultiViewMoSeg","url":"https://github.com/alex-xun-xu/MultiViewMoSeg"}],"syntology":null}},{"leaderboard":"/sota/motion-segmentation-on-kt3dmoseg","slug":"motion-segmentation-on-kt3dmoseg","dataset":"KT3DMoSeg","dataset_url":"/dataset/kt3dmoseg","rows_in_archive":1,"metrics":["Error"],"first_row_in_archive_order":{"model":"MultiViewClustering","paper_title":"Motion Segmentation by Exploiting Complementary Geometric Models","paper_url":"/paper/motion-segmentation-by-exploiting","paper_date":"2018-04-06","arxiv_id":"1804.02142","code_links":[{"title":"alex-xun-xu/MultiViewMoSeg","url":"https://github.com/alex-xun-xu/MultiViewMoSeg"}],"syntology":null}},{"leaderboard":"/sota/motion-segmentation-on-mtpv62","slug":"motion-segmentation-on-mtpv62","dataset":"MTPV62","dataset_url":null,"rows_in_archive":1,"metrics":["Classification Error"],"first_row_in_archive_order":{"model":"MVC","paper_title":"Motion Segmentation by Exploiting Complementary Geometric Models","paper_url":"/paper/motion-segmentation-by-exploiting","paper_date":"2018-04-06","arxiv_id":"1804.02142","code_links":[{"title":"alex-xun-xu/MultiViewMoSeg","url":"https://github.com/alex-xun-xu/MultiViewMoSeg"}],"syntology":null}}],"datasets":[{"url":"/dataset/hopkins155","name":"Hopkins155","full_name":"","num_papers_in_archive":92},{"url":"/dataset/apolloscape-1","name":"ApolloScape","full_name":"","num_papers_in_archive":74},{"url":"/dataset/hoi4d","name":"HOI4D","full_name":"","num_papers_in_archive":23},{"url":"/dataset/ev-imo","name":"EV-IMO","full_name":"","num_papers_in_archive":22},{"url":"/dataset/mod","name":"MOD++","full_name":"","num_papers_in_archive":3},{"url":"/dataset/kitti-15-msplus","name":"KITTI'15 MSplus","full_name":"KITTI'15 Instance Motion Segmentation - extension","num_papers_in_archive":1},{"url":"/dataset/kt3dmoseg","name":"KT3DMoSeg","full_name":"","num_papers_in_archive":1}],"subtasks":[],"parent_tasks":[],"papers":{"order":"repositories listed in the archive (desc), then date (desc); the archive holds no stars","population":"papers tagged with this task that list at least one repository in the archive","shown":30,"of":62,"tagged_in_all":212,"items":[{"url":"/paper/flownet3d-learning-scene-flow-in-3d-point","title":"FlowNet3D: Learning Scene Flow in 3D Point Clouds","date":"2018-06-04","arxiv_id":"1806.01411","repositories_listed":10,"syntology":{"n":7,"n_ran":2,"n_unverified":5,"n_pointer_only":0}},{"url":"/paper/sparse-subspace-clustering-algorithm-theory","title":"Sparse Subspace Clustering: Algorithm, Theory, and Applications","date":"2012-03-05","arxiv_id":"1203.1005","repositories_listed":4,"syntology":null},{"url":"/paper/exploiting-geometric-constraints-on-dense","title":"EpO-Net: Exploiting Geometric Constraints on Dense Trajectories for Motion Saliency","date":"2019-09-29","arxiv_id":"1909.13258","repositories_listed":3,"syntology":null},{"url":"/paper/self-supervised-scene-flow-estimation-with-4d","title":"Self-Supervised Scene Flow Estimation with 4-D Automotive Radar","date":"2022-03-02","arxiv_id":"2203.01137","repositories_listed":2,"syntology":{"n":12,"n_ran":2,"n_unverified":10,"n_pointer_only":0}},{"url":"/paper/moving-objects-detection-with-a-moving-camera","title":"Moving Objects Detection with a Moving Camera: A Comprehensive Review","date":"2020-01-15","arxiv_id":"2001.05238","repositories_listed":2,"syntology":null},{"url":"/paper/progressive-x-efficient-anytime-multi-model","title":"Progressive-X: Efficient, Anytime, Multi-Model Fitting Algorithm","date":"2019-06-05","arxiv_id":"1906.02290","repositories_listed":2,"syntology":null},{"url":"/paper/kdmos-knowledge-distillation-for-motion","title":"KDMOS:Knowledge Distillation for Motion Segmentation","date":"2025-06-17","arxiv_id":"2506.14130","repositories_listed":1,"syntology":null},{"url":"/paper/freegave-3d-physics-learning-from-dynamic-1","title":"FreeGave: 3D Physics Learning from Dynamic Videos by Gaussian Velocity","date":"2025-06-09","arxiv_id":"2506.07865","repositories_listed":1,"syntology":{"n":8,"n_ran":7,"n_unverified":1,"n_pointer_only":8}},{"url":"/paper/iterative-event-based-motion-segmentation-by","title":"Iterative Event-based Motion Segmentation by Variational Contrast Maximization","date":"2025-04-25","arxiv_id":"2504.18447","repositories_listed":1,"syntology":null},{"url":"/paper/wandering-around-a-bioinspired-approach-to","title":"Wandering around: A bioinspired approach to visual attention through object motion sensitivity","date":"2025-02-10","arxiv_id":"2502.06747","repositories_listed":1,"syntology":null},{"url":"/paper/cv-mos-a-cross-view-model-for-motion","title":"CV-MOS: A Cross-View Model for Motion Segmentation","date":"2024-08-25","arxiv_id":"2408.13790","repositories_listed":1,"syntology":null},{"url":"/paper/unsupervised-motion-segmentation-for","title":"Motion Segmentation for Neuromorphic Aerial Surveillance","date":"2024-05-24","arxiv_id":"2405.15209","repositories_listed":1,"syntology":null},{"url":"/paper/learning-to-detect-and-segment-mobile-objects","title":"MOD-UV: Learning Mobile Object Detectors from Unlabeled Videos","date":"2024-05-23","arxiv_id":"2405.14841","repositories_listed":1,"syntology":{"n":7,"n_ran":7,"n_unverified":0,"n_pointer_only":0}},{"url":"/paper/moving-object-segmentation-all-you-need-is","title":"Moving Object Segmentation: All You Need Is SAM (and Flow)","date":"2024-04-18","arxiv_id":"2404.12389","repositories_listed":1,"syntology":null},{"url":"/paper/motion2language-unsupervised-learning-of","title":"Motion2Language, unsupervised learning of synchronized semantic motion segmentation","date":"2023-10-16","arxiv_id":"2310.10594","repositories_listed":1,"syntology":null},{"url":"/paper/moving-object-detection-and-tracking-with-4d","title":"RaTrack: Moving Object Detection and Tracking with 4D Radar Point Cloud","date":"2023-09-18","arxiv_id":"2309.09737","repositories_listed":1,"syntology":null},{"url":"/paper/a-multi-scale-recurrent-framework-for-motion","title":"A Multi-Scale Recurrent Framework for Motion Segmentation With Event Camera","date":"2023-07-28","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/bootstrapping-objectness-from-videos-by","title":"Bootstrapping Objectness from Videos by Relaxed Common Fate and Visual Grouping","date":"2023-04-17","arxiv_id":"2304.08025","repositories_listed":1,"syntology":null},{"url":"/paper/hidden-gems-4d-radar-scene-flow-learning","title":"Hidden Gems: 4D Radar Scene Flow Learning Using Cross-Modal Supervision","date":"2023-03-01","arxiv_id":"2303.00462","repositories_listed":1,"syntology":{"n":1,"n_ran":0,"n_unverified":1,"n_pointer_only":0}},{"url":"/paper/unsupervised-space-time-network-for","title":"Unsupervised Space-Time Network for Temporally-Consistent Segmentation of Multiple Motions","date":"2023-01-01","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/gma3d-local-global-attention-learning-to","title":"GMA3D: Local-Global Attention Learning to Estimate Occluded Motions of Scene Flow","date":"2022-10-07","arxiv_id":"2210.03296","repositories_listed":1,"syntology":null},{"url":"/paper/dytanvo-joint-refinement-of-visual-odometry","title":"DytanVO: Joint Refinement of Visual Odometry and Motion Segmentation in Dynamic Environments","date":"2022-09-17","arxiv_id":"2209.08430","repositories_listed":1,"syntology":{"n":16,"n_ran":3,"n_unverified":13,"n_pointer_only":0}},{"url":"/paper/particlesfm-exploiting-dense-point","title":"ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wild","date":"2022-07-19","arxiv_id":"2207.09137","repositories_listed":1,"syntology":{"n":5,"n_ran":3,"n_unverified":2,"n_pointer_only":5}},{"url":"/paper/segmenting-moving-objects-via-an-object","title":"Segmenting Moving Objects via an Object-Centric Layered Representation","date":"2022-07-05","arxiv_id":"2207.02206","repositories_listed":1,"syntology":{"n":21,"n_ran":16,"n_unverified":5,"n_pointer_only":0}},{"url":"/paper/discovering-objects-that-can-move","title":"Discovering Objects that Can Move","date":"2022-03-18","arxiv_id":"2203.10159","repositories_listed":1,"syntology":null},{"url":"/paper/hoi4d-a-4d-egocentric-dataset-for-category","title":"HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction","date":"2022-03-03","arxiv_id":"2203.01577","repositories_listed":1,"syntology":{"n":5,"n_ran":0,"n_unverified":5,"n_pointer_only":0}},{"url":"/paper/em-driven-unsupervised-learning-for-efficient","title":"EM-driven unsupervised learning for efficient motion segmentation","date":"2022-01-06","arxiv_id":"2201.02074","repositories_listed":1,"syntology":null},{"url":"/paper/image-reconstruction-from-events-why-learn-it","title":"Formulating Event-based Image Reconstruction as a Linear Inverse Problem with Deep Regularization using Optical Flow","date":"2021-12-12","arxiv_id":"2112.06242","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_unverified":0,"n_pointer_only":0}},{"url":"/paper/monocular-arbitrary-moving-object-discovery","title":"Monocular Arbitrary Moving Object Discovery and Segmentation","date":"2021-11-22","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/event-based-motion-segmentation-by-cascaded","title":"Event-based Motion Segmentation by Cascaded Two-Level Multi-Model Fitting","date":"2021-11-05","arxiv_id":"2111.03483","repositories_listed":1,"syntology":null}],"syntology_records":10,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; our fixture could not drive it","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"}},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"}}