mlsanityData to 7 Oct 2026

Robust Online Video Instance Segmentation with Track Queries

arXiv 2211.09108 ↗First posted 16 Nov 2022cs.CVcited by 8

Not on the map. Why?

Abstract (arXiv)

Recently, transformer-based methods have achieved impressive results on Video Instance Segmentation (VIS). However, most of these top-performing methods run in an offline manner by processing the entire video clip at once to predict instance mask volumes. This makes them incapable of handling the long videos that appear in challenging new video instance segmentation datasets like UVO and OVIS. We propose a fully online transformer-based video instance segmentation model that performs comparably to top offline methods on the YouTube-VIS 2019 benchmark and considerably outperforms them on UVO and OVIS. This method, called Robust Online Video Segmentation (ROVIS), augments the Mask2Former image instance segmentation model with track queries, a lightweight mechanism for carrying track information from frame to frame, originally introduced by the TrackFormer method for multi-object tracking. We show that, when combined with a strong enough image segmentation architecture, track queries can exhibit impressive accuracy while not being constrained to short videos.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 48 of its 63 references that are in the map. First posted on arXiv 16 Nov 2022.

Builds on

48 of 63 references.

References · 48

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  4. 2020End-to-End Object Detection with Transformerscited by 7,384
  5. 2017Mask R-CNNcited by 7,259
  6. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,338
  7. 2016V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentationcited by 2,284
  8. 2017The Kinetics Human Action Video Datasetcited by 1,894
  9. 2019MMDetection: Open MMLab Detection Toolbox and Benchmarkcited by 1,379
  10. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,282
  11. 2021Per-Pixel Classification is Not All You Need for Semantic Segmentationcited by 1,148
  12. 2021YOLOX: Exceeding YOLO Series in 2021cited by 1,064
36 more
  1. 2017Simple Online and Realtime Tracking with a Deep Association Metriccited by 908
  2. 2016Simple Online and Realtime Trackingcited by 862
  3. 2021ByteTrack: Multi-Object Tracking by Associating Every Detection Boxcited by 812
  4. 2016MOT16: A Benchmark for Multi-Object Trackingcited by 531
  5. 2021TrackFormer: Multi-Object Tracking with Transformerscited by 515
  6. 2020FairMOT: On the Fairness of Detection and Re-Identification in Multiple Object Trackingcited by 489
  7. 2020Tracking Objects as Pointscited by 483
  8. 2020End-to-End Video Instance Segmentation with Transformerscited by 443
  9. 2019Video Instance Segmentationcited by 391
  10. 2019Tracking without bells and whistlescited by 382
  11. 2021MOTR: End-to-End Multiple-Object Tracking with Transformercited by 379
  12. 2020TransTrack: Multiple Object Tracking with Transformercited by 370
  13. 2020MOT20: A benchmark for multi object tracking in crowded scenescited by 355
  14. 2019Towards Real-Time Multi-Object Trackingcited by 311
  15. 2019MOTS: Multi-Object Tracking and Segmentationcited by 222
  16. 2020Quasi-Dense Similarity Learning for Multiple Object Trackingcited by 220
  17. 2019Learning a Neural Solver for Multiple Object Trackingcited by 198
  18. 2021Mask2Former for Video Instance Segmentationcited by 169
  19. 2021Instances as Queriescited by 157
  20. 2021Occluded Video Instance Segmentation: A Benchmarkcited by 154
  21. 2020TAO: A Large-Scale Benchmark for Tracking Any Objectcited by 146
  22. 2021Unidentified Video Objects: A Benchmark for Dense, Open-World Segmentationcited by 133
  23. 2017Detect to Track and Track to Detectcited by 131
  24. 2019Classifying, Segmenting, and Tracking Object Instances in Video with Mask Propagationcited by 125
  25. 2020STEm-Seg: Spatio-temporal Embeddings for Instance Segmentation in Videoscited by 121
  26. 2020SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentationcited by 108
  27. 2021Video Instance Segmentation using Inter-Frame Communication Transformerscited by 105
  28. 2022In Defense of Online Models for Video Instance Segmentationcited by 97
  29. 2022MinVIS: A Minimal Video Instance Segmentation Framework without Video-based Trainingcited by 94
  30. 2021Crossover Learning for Fast Online Video Instance Segmentationcited by 83
  31. 2019Spatial-Temporal Relation Networks for Multi-Object Trackingcited by 74
  32. 2021SiamMOT: Siamese Multi-Object Trackingcited by 62
  33. 2021Opening up Open-World Trackingcited by 50
  34. 2020Learning Video Instance Segmentation with Recurrent Graph Neural Networkscited by 6
  35. 2021Multi-Object Tracking with Hallucinated and Unlabeled Videoscited by 6
  36. 2021UVO Challenge on Video-based Open-World Segmentation 2021: 1st Place Solutioncited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 4; 2024: 3; 2025: 0; 2026: 1.

8 papers.

  1. 2023Segment Anything Meets Point Trackingcited by 81
  2. 2023DVIS: Decoupled Video Instance Segmentation Frameworkcited by 77
  3. 2023TCOVIS: Temporally Consistent Online Video Instance Segmentationcited by 18
  4. 2024Computer Vision for Primate Behavior Analysis in the Wildcited by 10
  5. 2023GRAtt-VIS: Gated Residual Attention for Auto Rectifying Video Instance Segmentationcited by 9
  6. 2024SAM-PD: How Far Can SAM Take Us in Tracking and Segmenting Anything in Videos by Prompt Denoisingcited by 4
  7. 2024Instance Brownian Bridge as Texts for Open-vocabulary Video Instance Segmentationcited by 3
  8. 2026SA-VIS: Sparse frame Annotations for training Video Instance Segmentationcited by 0