mlsanityData to 14 Oct 2026

Self-Supervised Video Object Segmentation by Motion-Aware Mask Propagation

arXiv 2107.12569 ↗First posted 27 Jul 2021cs.CVcited by 11

Not on the map. Why?

Abstract (arXiv)

We propose a self-supervised spatio-temporal matching method, coined Motion-Aware Mask Propagation (MAMP), for video object segmentation. MAMP leverages the frame reconstruction task for training without the need for annotations. During inference, MAMP extracts high-resolution features from each frame to build a memory bank from the features as well as the predicted masks of selected past frames. MAMP then propagates the masks from the memory bank to subsequent frames according to our proposed motion-aware spatio-temporal matching module to handle fast motion and long-term matching scenarios. Evaluation on DAVIS-2017 and YouTube-VOS datasets show that MAMP achieves state-of-the-art performance with stronger generalization ability compared to existing self-supervised methods, i.e., 4.2% higher mean J&F on DAVIS-2017 and 4.85% higher mean J&F on the unseen categories of YouTube-VOS than the nearest competitor. Moreover, MAMP performs at par with many supervised video object segmentation methods. Our code is available at: https://github.com/bo-miao/MAMP.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Builds on 40 of its 47 references that are in the map. First posted on arXiv 27 Jul 2021.

Builds on

40 of 47 references.

References · 40

  1. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,057
  2. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,415
  3. 2017The 2017 DAVIS Challenge on Video Object Segmentationcited by 957
  4. 2019Video Object Segmentation using Space-Time Memory Networkscited by 438
  5. 2017Future Frame Prediction for Anomaly Detection -- A New Baselinecited by 391
  6. 2018YouTube-VOS: A Large-Scale Video Object Segmentation Benchmarkcited by 383
  7. 2014Memory Networkscited by 363
  8. 2018Fast Online Object Tracking and Segmentation: A Unifying Approachcited by 272
  9. 2019Learning Correspondence from the Cycle-Consistency of Timecited by 269
  10. 2016One-Shot Video Object Segmentationcited by 235
  11. 2019FEELVOS: Fast End-to-End Embedding Learning for Video Object Segmentationcited by 178
  12. 2019Video Representation Learning by Dense Predictive Codingcited by 175
28 more
  1. 2018Tracking Emerges by Colorizing Videoscited by 166
  2. 2021VideoMoCo: Contrastive Video Representation Learning with Temporally Adversarial Examplescited by 154
  3. 2016Learning Video Object Segmentation from Static Imagescited by 151
  4. 2018Self-Supervised Video Representation Learning with Space-Time Cubic Puzzlescited by 149
  5. 2020Collaborative Video Object Segmentation by Foreground-Background Integrationcited by 142
  6. 2020Self-supervised Video Representation Learning by Pace Predictioncited by 139
  7. 2020Kernelized Memory Network for Video Object Segmentationcited by 111
  8. 2017Online Adaptation of Convolutional Neural Networks for Video Object Segmentationcited by 109
  9. 2020MAST: A Memory-Augmented Self-supervised Trackercited by 109
  10. 2020Video Object Segmentation with Adaptive Feature Bank and Uncertain-Region Refinementcited by 101
  11. 2018Efficient Video Object Segmentation via Network Modulationcited by 98
  12. 2017Video Object Segmentation Without Temporal Informationcited by 93
  13. 2019Self-supervised Spatio-temporal Representation Learning for Videos by Predicting Motion and Appearance Statisticscited by 92
  14. 2018VideoMatch: Matching based Video Object Segmentationcited by 90
  15. 2019RVOS: End-to-End Recurrent Network for Video Object Segmentationcited by 88
  16. 2021Efficient Regional Memory Network for Video Object Segmentationcited by 86
  17. 2018PReMVOS: Proposal-generation, Refinement and Merging for Video Object Segmentationcited by 81
  18. 2019Joint-task Self-supervised Learning for Temporal Correspondencecited by 80
  19. 2018Blazingly Fast Video Object Segmentation with Pixel-Wise Metric Learningcited by 74
  20. 2018A Generative Appearance Model for End-to-end Video Object Segmentationcited by 66
  21. 2020Fast Video Object Segmentation using the Global Context Modulecited by 65
  22. 2019RANet: Ranking Attention Network for Fast Video Object Segmentationcited by 60
  23. 2018Video Object Segmentation with Joint Re-identification and Attention-Aware Mask Propagationcited by 56
  24. 2020Self-supervised Video Representation Learning Using Inter-intra Contrastive Frameworkcited by 46
  25. 2020Learning Video Object Segmentation from Unlabeled Videoscited by 42
  26. 2020Fast Video Object Segmentation With Temporal Aggregation Network and Dynamic Template Matchingcited by 37
  27. 2020Make One-Shot Video Object Segmentation Efficient Againcited by 30
  28. 2019RPM-Net: Robust Pixel-Level Matching Networks for Self-Supervised Video Object Segmentationcited by 3

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 3; 2023: 3; 2024: 3; 2025: 1.

11 papers.

  1. 2022Segmenting Moving Objects via an Object-Centric Layered Representationcited by 63
  2. 2023Spectrum-guided Multi-granularity Referring Video Object Segmentationcited by 48
  3. 2024Moving Object Segmentation: All You Need Is SAM (and Flow)cited by 30
  4. 2021HODOR: High-level Object Descriptors for Object Re-segmentation in Video Learned from Static Imagescited by 21
  5. 2022Region Aware Video Object Segmentation with Deep Motion Modelingcited by 19
  6. 2024Temporally Consistent Referring Video Object Segmentation with Hybrid Memorycited by 14
  7. 2022Spatial-then-Temporal Self-Supervised Learning for Video Correspondencecited by 13
  8. 2023Appearance-Based Refinement for Object-Centric Motion Segmentationcited by 11
  9. 2024Referring Human Pose and Mask Estimation in the Wildcited by 6
  10. 2023Learning Fine-Grained Features for Pixel-wise Video Correspondencescited by 4
  11. 2025Learning Appearance and Motion Cues for Panoptic Trackingcited by 1