pith. sign in

arxiv: 2606.20140 · v2 · pith:ENWDJZJ2new · submitted 2026-06-18 · 💻 cs.CV

SA-VIS: Sparse frame Annotations for training Video Instance Segmentation

Edoardo Mello Rella , Ajad Chhatkuli , Shipra Jain , Ender Konukoglu , Luc Van Gool This is my paper

Pith reviewed 2026-06-30 10:43 UTC · model grok-4.3

classification 💻 cs.CV
keywords video instance segmentationsparse annotationsfeature propagationonline VISYouTube-VISOVISinstance queries
0
0 comments X

The pith

A simple feature propagation module trains video instance segmentation on one-fifth the labels with only a 0.4 percent accuracy drop.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that video instance segmentation does not need dense per-frame labels during training. Its Past-frames Feature Propagation module pulls low-dimensional features from sparsely labeled past frames to track instance evolution and resolve ambiguities. Combined with light frame-specific queries, the resulting SA-VIS model closes most of the performance gap to dense-annotation training. The approach yields large gains over its own baseline on standard benchmarks while cutting annotation cost dramatically.

Core claim

The Past-frames Feature Propagation module aggregates low-dimensional features across multiple past frames inside the image encoder, supplying temporal context that lets the model learn instance motion and identity from sparse video labels. When paired with frame-specific Instance Queries, this yields end-to-end training whose accuracy on YouTube-VIS and OVIS falls only 0.4 percent below the same architecture trained on fully dense annotations.

What carries the argument

Past-frames Feature Propagation (PFP) module, which aggregates low-dimensional features from the image encoder of multiple frames to supply temporal context for instance modeling.

If this is right

  • SA-VIS raises accuracy over its baseline on YouTube-VIS 2019/2021/2022 and on OVIS.
  • In low-annotation regimes the method improves AP by more than 1 percent over prior state-of-the-art.
  • The same architecture trained on 20 percent of frames retains nearly all accuracy of its dense counterpart.
  • No additional temporal modeling components are required once the propagation module is added.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The result implies that many video tasks may tolerate far sparser supervision once simple feature reuse is introduced.
  • Extending the propagation window to longer sequences could test whether the low-dimensional cue remains sufficient.
  • The same module might reduce labeling needs in related tasks such as video object tracking.

Load-bearing premise

Low-dimensional features from sparsely labeled past frames are enough to capture how instances move and stay distinct without dense labels or extra temporal networks.

What would settle it

Train SA-VIS on 1/5 annotations and measure AP on YouTube-VIS 2019; if the drop versus the dense version exceeds 2 percent, the core claim fails.

Figures

Figures reproduced from arXiv: 2606.20140 by Ajad Chhatkuli, Edoardo Mello Rella, Ender Konukoglu, Luc Van Gool, Shipra Jain.

Figure 1
Figure 1. Figure 1: SA-VIS: we propose a method that includes past frames awareness and generates frame￾specific instance queries for the task of video instance segmentation. Thanks to the addition of the shown modules, we propose a method that i) generates queries based on the objects actually visible in the image (FSI Queries), and ii) uses the past frames to build useful and lightweight contextual information (PFP). The pr… view at source ↗
Figure 2
Figure 2. Figure 2: Qualitative results: Visualization of SA-VIS on a set of challenging scenes. 4.3 Qualitative Results In [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
read the original abstract

Recent online video instance segmentation (VIS) methods have achieved impressive results, thus becoming the preferred approach to segment instances in videos. Despite the resurgence of impressive single image models, the online (or semi-online) VIS approaches outperform single-image models (e.g., based on SAM) by using long sequences of densely annotated frames during training. However,such a training setup of VIS is expensive in the sense of compute as well as dense annotations required. In order to solve these major flaws, we argue that the effective modeling of the instances and their evolution in videos do not require densely annotated frames. To that end, we propose a simple and effective module, called Past-frames Feature Propagation (PFP) which aggregates low-dimensional features from the image encoder of multiple frames. This simple low-compute module provides tremendous learning capability in using sparse video frame labels for end-to-end training. Combined with a light-weight frame-specific Instance Queries, our Sparse frame Annotation VIS (SA-VIS) significantly improves performance over its baseline. Most interestingly, our simple design that avoids complexities effectively bridges the gap in accuracy between training on sparsely and densely annotated video sequences. This translates to a mere 0.4% drop in performance of SA-VIS when using annotations for only 1/5 of the images in the dataset. Empirically, SA-VIS shows strong improvements over the baseline on YouTube-VIS 2019/2021/2022 and Occluded VIS (OVIS) and an over 1% improvement in AP on the state-of-the-art in a limited annotations scenario.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes SA-VIS for online video instance segmentation, introducing a Past-frames Feature Propagation (PFP) module that aggregates low-dimensional features from the image encoder across multiple (sparsely annotated) frames, combined with lightweight frame-specific instance queries. The central empirical claim is that this simple design bridges the performance gap between sparse and dense annotation regimes, yielding only a 0.4% AP drop when training on annotations for 1/5 of the frames versus full dense supervision, while also outperforming baselines and prior SOTA by over 1% AP in the limited-annotation setting on YouTube-VIS 2019/2021/2022 and OVIS.

Significance. If the reported numbers hold after verification, the result would be significant for reducing the annotation and compute burden of VIS training. The approach demonstrates that low-dimensional feature aggregation plus per-frame queries can nearly close the dense-vs-sparse gap without extra temporal modules. Credit is given for evaluating on multiple standard benchmarks (YouTube-VIS 2019/2021/2022 and OVIS) and for the reproducible experimental setup implied by the benchmark comparisons.

major comments (2)
  1. [Abstract] Abstract: the central claim of a 'mere 0.4% drop' when using annotations for only 1/5 of the images is presented without error bars, number of runs, statistical significance tests, or explicit baseline definitions (dense vs. sparse SA-VIS), which is load-bearing for the assertion that the gap is effectively bridged.
  2. [§4] §4 (Experiments): the reported gains over baselines and SOTA in the sparse regime lack details on ablation controls isolating PFP from the frame-specific queries, as well as variance across runs, undermining verification of the 0.4% claim and the 'simple design' sufficiency argument.
minor comments (2)
  1. [Abstract] Abstract: missing space after the comma in 'However,such a training setup'.
  2. [Abstract] Abstract: 'an over 1% improvement' appears to be a typographical error and should read 'and over 1% improvement'.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on clarifying the empirical claims. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim of a 'mere 0.4% drop' when using annotations for only 1/5 of the images is presented without error bars, number of runs, statistical significance tests, or explicit baseline definitions (dense vs. sparse SA-VIS), which is load-bearing for the assertion that the gap is effectively bridged.

    Authors: We agree the abstract would benefit from explicit context. In revision we will define the dense vs. sparse SA-VIS comparison (same architecture and training protocol, differing only in annotation density) and report error bars plus number of runs. The 0.4% figure is the direct AP difference on YouTube-VIS between full dense supervision and 1/5-frame annotations for the identical SA-VIS model. revision: yes

  2. Referee: [§4] §4 (Experiments): the reported gains over baselines and SOTA in the sparse regime lack details on ablation controls isolating PFP from the frame-specific queries, as well as variance across runs, undermining verification of the 0.4% claim and the 'simple design' sufficiency argument.

    Authors: We will add the requested ablation controls that isolate PFP from frame-specific queries and will report standard deviation across multiple runs in the revised §4. These additions will allow direct verification of the 0.4% gap and the contribution of each component. revision: yes

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper proposes an empirical architecture (PFP module for aggregating low-dimensional encoder features plus frame-specific queries) and reports benchmark results on YouTube-VIS and OVIS. No derivation chain, equations, or first-principles claims exist that reduce by construction to fitted inputs or self-citations. Performance numbers (e.g., 0.4% AP drop on 1/5 annotations) are external comparisons against baselines and prior SOTA; the central claim is falsifiable on held-out data and does not rely on self-referential definitions or imported uniqueness theorems.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the design appears to reuse standard image-encoder features and instance-query mechanisms.

pith-pipeline@v0.9.1-grok · 5833 in / 1068 out tokens · 29263 ms · 2026-06-30T10:43:00.198861+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

51 extracted references · 6 canonical work pages

  1. [1]

    In: Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16

    Cao, J., Anwer, R.M., Cholakkal, H., Khan, F.S., Pang, Y ., Shao, L.: Sipmask: Spatial information preservation for fast image and video instance segmentation. In: Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16. pp. 1–18. Springer (2020)

    2020

  2. [2]

    In: European conference on computer vision

    Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., Zagoruyko, S.: End-to-end object detection with transformers. In: European conference on computer vision. pp. 213–229. Springer (2020)

    2020

  3. [3]

    ArXivabs/2112.10764(2021), https://api

    Cheng, B., Choudhuri, A., Misra, I., Kirillov, A., Girdhar, R., Schwing, A.G.: Mask2former for video instance segmentation. ArXivabs/2112.10764(2021), https://api. semanticscholar.org/CorpusID:245335013

    work page arXiv 2021

  4. [4]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2020)

    Cheng, B., Collins, M.D., Zhu, Y ., Liu, T., Huang, T.S., Adam, H., Chen, L.C.: Panoptic-deeplab: A simple, strong, and fast baseline for bottom-up panoptic segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2020)

    2020

  5. [5]

    Cheng, B., Misra, I., Schwing, A.G., Kirillov, A., Girdhar, R.: Masked-attention mask trans- former for universal image segmentation (2022)

    2022

  6. [6]

    Cheng, B., Schwing, A.G., Kirillov, A.: Per-pixel classification is not all you need for semantic segmentation (2021)

    2021

  7. [7]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Choudhuri, A., Chowdhary, G., Schwing, A.G.: Context-aware relative object queries to unify video instance and panoptic segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 6377–6386 (June 2023)

    2023

  8. [8]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Choudhuri, A., Chowdhary, G., Schwing, A.G.: Context-aware relative object queries to unify video instance and panoptic segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 6377–6386 (2023)

    2023

  9. [9]

    IEEE Transactions on Circuits and Systems for Video Technology pp

    Fang, H., Zhang, T., Zhou, X., Zhang, X.: Learning better video query with sam for video instance segmentation. IEEE Transactions on Circuits and Systems for Video Technology pp. 1–1 (2024). https://doi.org/10.1109/TCSVT.2024.3361076

    work page doi:10.1109/tcsvt.2024.3361076 2024

  10. [10]

    Fischer, T., Huang, T.E., Pang, J., Qiu, L., Chen, H., Darrell, T., Yu, F.: Qdtrack: Quasi-dense similarity learning for appearance-only multiple object tracking (2023)

    2023

  11. [11]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Fu, Y ., Yang, L., Liu, D., Huang, T.S., Shi, H.: Compfeat: Comprehensive feature aggrega- tion for video instance segmentation. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 35, pp. 1361–1369 (2021)

    2021

  12. [12]

    In: Proceedings of the IEEE International Conference on Computer Vision

    Gadde, R., Jampani, V ., Gehler, P.V .: Semantic video cnns through representation warping. In: Proceedings of the IEEE International Conference on Computer Vision. pp. 4453–4462 (2017)

    2017

  13. [13]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Han, S.H., Hwang, S., Oh, S.W., Park, Y ., Kim, H., Kim, M.J., Kim, S.J.: Visolo: Grid-based space-time aggregation for efficient online video instance segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2896–2905 (2022)

    2022

  14. [14]

    arXiv preprint arXiv:2305.17096 (2023)

    Hannan, T., Koner, R., Bernhard, M., Shit, S., Menze, B., Tresp, V ., Schubert, M., Seidl, T.: Gratt-vis: Gated residual attention for auto rectifying video instance segmentation. arXiv preprint arXiv:2305.17096 (2023)

    work page arXiv 2023

  15. [15]

    Advances in Neural Information Processing Systems35, 19370–19383 (2022)

    He, F., Zhang, H., Gao, N., Jia, J., Shan, Y ., Zhao, X., Huang, K.: Inspro: Propagating instance query and proposal for online video instance segmentation. Advances in Neural Information Processing Systems35, 19370–19383 (2022)

    2022

  16. [16]

    In: Proceedings of the IEEE International Conference on Computer Vision (ICCV) (Oct 2017)

    He, K., Gkioxari, G., Dollar, P., Girshick, R.: Mask r-cnn. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV) (Oct 2017)

    2017

  17. [17]

    In: Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (June 2016)

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (June 2016)

    2016

  18. [18]

    In: CVPR (2023)

    Heo, M., Hwang, S., Hyun, J., Kim, H., Oh, S.W., Lee, J.Y ., Kim, S.J.: A generalized framework for video instance segmentation. In: CVPR (2023)

    2023

  19. [19]

    In: Advances in Neural Information Processing Systems (2022) 10

    Heo, M., Hwang, S., Oh, S.W., Lee, J.Y ., Kim, S.J.: Vita: Video instance segmentation via object token association. In: Advances in Neural Information Processing Systems (2022) 10

    2022

  20. [20]

    Huang, D.A., Yu, Z., Anandkumar, A.: Minvis: A minimal video instance segmentation framework without video-based training (2022)

    2022

  21. [21]

    Advances in Neural Information Processing Systems34, 13352– 13363 (2021)

    Hwang, S., Heo, M., Oh, S.W., Kim, S.J.: Video instance segmentation using inter-frame communication transformers. Advances in Neural Information Processing Systems34, 13352– 13363 (2021)

    2021

  22. [22]

    In: Computer Vision – ECCV 2022 Workshops: Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part IV (2023)

    Jiang, Z., Gu, Z., Peng, J., Zhou, H., Liu, L., Wang, Y ., Tai, Y ., Wang, C., Zhang, L.: Stc: Spatio-temporal contrastive learning for video instance segmentation. In: Computer Vision – ECCV 2022 Workshops: Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part IV (2023)

    2022

  23. [23]

    In: Proceedings of the IEEE International Conference on Computer Vision

    Jin, X., Li, X., Xiao, H., Shen, X., Lin, Z., Yang, J., Chen, Y ., Dong, J., Liu, L., Jie, Z., et al.: Video scene parsing with predictive feature learning. In: Proceedings of the IEEE International Conference on Computer Vision. pp. 5580–5588 (2017)

    2017

  24. [24]

    In: CVPR (2023)

    Ke, L., Danelljan, M., Ding, H., Tai, Y .W., Tang, C.K., Yu, F.: Mask-free video instance segmentation. In: CVPR (2023)

    2023

  25. [25]

    arXiv preprint arXiv:2208.10547 (2022)

    Koner, R., Hannan, T., Shit, S., Sharifzadeh, S., Schubert, M., Seidl, T., Tresp, V .: Instance- former: An online video instance segmentation framework. arXiv preprint arXiv:2208.10547 (2022)

    work page arXiv 2022

  26. [26]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Lin, C.C., Hung, Y ., Feris, R., He, L.: Video instance segmentation tracking with a modified vae architecture. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 13147–13157 (2020)

    2020

  27. [27]

    In: Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13

    Lin, T.Y ., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., Zitnick, C.L.: Microsoft coco: Common objects in context. In: Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13. pp. 740–755. Springer (2014)

    2014

  28. [28]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Liu, D., Cui, Y ., Tan, W., Chen, Y .: Sg-net: Spatial granularity network for one-stage video instance segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 9816–9825 (2021)

    2021

  29. [29]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Liu, Z., Lin, Y ., Cao, Y ., Hu, H., Wei, Y ., Zhang, Z., Lin, S., Guo, B.: Swin transformer: Hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 10012–10022 (2021)

    2021

  30. [30]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Nilsson, D., Sminchisescu, C.: Semantic video segmentation by gated recurrent flow propaga- tion. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 6819–6828 (2018)

    2018

  31. [31]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Porzi, L., Hofinger, M., Ruiz, I., Serrat, J., Bulo, S.R., Kontschieder, P.: Learning multi-object tracking and segmentation from automatic annotations. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 6846–6855 (2020)

    2020

  32. [32]

    In: Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round

    Qi, J., Gao, Y ., Hu, Y ., Wang, X., Liu, X., Bai, X., Belongie, S., Yuille, A., Torr, P., Bai, S.: Occluded video instance segmentation: Dataset and ICCV 2021 challenge. In: Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round

    2021

  33. [33]

    (2021),https://openreview.net/forum?id=IfzTefIU_3j

    2021

  34. [34]

    In: Com- puter Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part I 16

    Tian, Z., Shen, C., Chen, H.: Conditional convolutions for instance segmentation. In: Com- puter Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part I 16. pp. 282–298. Springer (2020)

    2020

  35. [35]

    In: Guyon, I., Luxburg, U.V ., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, L.u., Polosukhin, I.: Attention is all you need. In: Guyon, I., Luxburg, U.V ., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R. (eds.) Advances in Neural Information Process- ing Systems. vol. 30. Curran Associates, Inc. (2017), https://proceedings.neur...

    2017

  36. [36]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2019)

    V oigtlaender, P., Krause, M., Osep, A., Luiten, J., Sekar, B.B.G., Geiger, A., Leibe, B.: Mots: Multi-object tracking and segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (June 2019)

    2019

  37. [37]

    In: Proc

    Wang, Y ., Xu, Z., Wang, X., Shen, C., Cheng, B., Shen, H., Xia, H.: End-to-end video instance segmentation with transformers. In: Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR) (2021) 11

    2021

  38. [38]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Wu, D., Wang, T., Zhang, Y ., Zhang, X., Shen, J.: Onlinerefer: A simple online baseline for referring video object segmentation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2761–2770 (2023)

    2023

  39. [39]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Wu, J., Cao, J., Song, L., Wang, Y ., Yang, M., Yuan, J.: Track to detect and segment: An online multi-object tracker. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 12352–12361 (2021)

    2021

  40. [40]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Wu, J., Yarram, S., Liang, H., Lan, T., Yuan, J., Eledath, J., Medioni, G.: Efficient video instance segmentation via tracklet query and proposal. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 959–968 (2022)

    2022

  41. [41]

    arXiv preprint arXiv:2112.08275 (2021)

    Wu, J., Jiang, Y ., Zhang, W., Bai, X., Bai, S.: Seqformer: a frustratingly simple model for video instance segmentation. arXiv preprint arXiv:2112.08275 (2021)

    work page arXiv 2021

  42. [42]

    In: ECCV (2022)

    Wu, J., Liu, Q., Jiang, Y ., Bai, S., Yuille, A., Bai, X.: In defense of online models for video instance segmentation. In: ECCV (2022)

    2022

  43. [43]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (October 2019)

    Yang, L., Fan, Y ., Xu, N.: Video instance segmentation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (October 2019)

    2019

  44. [44]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)

    Yang, S., Fang, Y ., Wang, X., Li, Y ., Fang, C., Shan, Y ., Feng, B., Liu, W.: Crossover learning for fast online video instance segmentation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 8043–8052 (October 2021)

    2021

  45. [45]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Yang, S., Wang, X., Li, Y ., Fang, Y ., Fang, J., Liu, W., Zhao, X., Shan, Y .: Temporally efficient vision transformer for video instance segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2885–2895 (2022)

    2022

  46. [46]

    Advances in Neural Information Processing Systems35, 36324–36336 (2022)

    Yang, Z., Yang, Y .: Decoupling features in hierarchical propagation for video object segmenta- tion. Advances in Neural Information Processing Systems35, 36324–36336 (2022)

    2022

  47. [47]

    Ying, K., Zhong, Q., Mao, W., Wang, Z., Chen, H., Wu, L.Y ., Liu, Y ., Fan, C., Zhuge, Y ., Shen, C.: CTVIS: Consistent Training for Online Video Instance Segmentation (2023)

    2023

  48. [48]

    arXiv preprint arXiv:2211.09108 (2022)

    Zhan, Z., McKee, D., Lazebnik, S.: Robust online video instance segmentation with track queries. arXiv preprint arXiv:2211.09108 (2022)

    work page arXiv 2022

  49. [49]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)

    Zhang, T., Tian, X., Wu, Y ., Ji, S., Wang, X., Zhang, Y ., Wan, P.: Dvis: Decoupled video instance segmentation framework. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 1282–1291 (October 2023)

    2023

  50. [50]

    In: MultiMedia Modeling: 26th International Conference, MMM 2020, Daejeon, South Korea, January 5–8, 2020, Proceedings, Part I 26

    Zhang, X., Han, G., He, W.: Unsupervised feature propagation for fast video object detection using generative adversarial networks. In: MultiMedia Modeling: 26th International Conference, MMM 2020, Daejeon, South Korea, January 5–8, 2020, Proceedings, Part I 26. pp. 617–627. Springer (2020)

    2020

  51. [51]

    In: International Conference on Learning Representations (2021),https://openreview.net/forum?id=gZ9hCDWe6ke 12

    Zhu, X., Su, W., Lu, L., Li, B., Wang, X., Dai, J.: Deformable {detr}: Deformable transformers for end-to-end object detection. In: International Conference on Learning Representations (2021),https://openreview.net/forum?id=gZ9hCDWe6ke 12

    2021