Pith. sign in

REVIEW 3 major objections 5 minor 52 references

A2VIS: Amodal-Aware Approach to Video Instance Segmentation

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A2VIS claims that predicting full object shapes, including occluded parts, makes video instance tracking more stable and cuts identity switches on occlusion-heavy benchmarks.

desk verdict Solid but narrower than claimed: the gains are real for partial in-frame occlusion, not for full disappearance/reappearance. read the letter →

arxiv 2412.01147 v2 pith:GK2YZIZ3 submitted 2024-12-02 cs.CV

Minh Tran , Thang Pham , Winston Bounsavy , Tri Nguyen , Ngan Le This is my paper · ORCID
classification cs.CV
keywords amodalsegmentationvideoinstancemultipleobjecttrackingocclusionhandlingglobalprototypespatiotemporalmaskedattentionFISHBOWLSAILVOS
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces A2VIS, a video instance segmentation framework that predicts, for each object, both its visible mask and its amodal mask, meaning the full shape including parts hidden by occlusion. The central claim is that feeding amodal masks into the tracking process makes instance identity more stable over time, because an object's full shape changes less abruptly than its visible outline during occlusion. A2VIS encodes this by compressing each object into a global instance prototype that accumulates amodal information across the whole video, and by using a dedicated head, SAMH, that draws on both nearby visible frames and long-range amodal knowledge. On FISHBOWL and SAIL VOS the authors report gains over prior VIS and amodal-VIS baselines, including higher IDF1 and fewer identity switches. If the claim holds, amodal awareness is a practical inductive bias for occlusion-robust tracking, not just a way to render completed masks.

What carries the argument

The load-bearing object is the Spatiotemporal-prior Amodal Mask Head (SAMH), a mask-decoding module that predicts amodal masks by attending to two spatiotemporal-prior masks: VSPM, built from visible segmentations inside the current clip, and ASPM, built from amodal segmentations accumulated in the global instance prototypes across clips. The attention operation is $\mathbf{p}^G_l = \mathrm{softmax}(T^k + QK^\top)V + \mathbf{p}^G_{l-1}$, where $T^k$ is the union of VSPM and ASPM, so the head can only look at locations where the instance is known to be visible locally or amodally globally. Alongside SAMH, the global instance prototypes act as a compressed, video-wide memory that keeps each object's identity and full-shape information, updated clip by clip through cross-attention. Together they carry the paper's argument that amodal information stabilizes tracking: SAMH supplies the full-shape predictions, and the prototypes propagate that information across time.

What would settle it

Run A2VIS and the GenVIS-style baselines on a benchmark where objects are frequently fully occluded by other objects or leave the frame entirely and then return, with identity annotations for reappearance; if A2VIS's identity-switch advantage disappears or reverses, the claim that amodal awareness stabilizes tracking beyond in-frame partial occlusion fails. A cheaper test is to add out-of-frame amodal annotations to FISHBOWL or SAIL VOS and check whether training on them changes the reported gaps.

Watch

Extended reading notes

Core claim

The paper's discovery candidate is that amodal segmentation is a useful representation for video instance tracking, not merely an output task. A2VIS treats the video as a sequence of clips, maintains a set of global instance prototypes that represent every object across the whole video, and decodes both visible and amodal masks from those prototypes. The amodal masks are produced by the Spatiotemporal-prior Amodal Mask Head, whose masked attention uses a visible spatiotemporal-prior mask (short-range, from visible segments in nearby frames) and an amodal spatiotemporal-prior mask (long-range, from accumulated amodal knowledge in the global prototypes). Empirically, the paper reports that A2VIS surpasses the prior GenVIS baseline on FISHBOWL and SAIL VOS across segmentation and bounding-box tracking metrics, and that the amodal-aware design reduces identity switches; ablation studies attribute the gain to the combination of the two spatiotemporal priors. The paper also introduces amodal extensions of VIS baselines and shows A2VIS ahead of them. The reported gains are confined to in-frame occlusions, because the amodal annotations in FISHBOWL and SAIL VOS clip masks to the frame and do not cover objects that disappear and reappear.

Load-bearing premise

The load-bearing assumption is that the amodal masks in the training data correctly represent full object shape under the occlusions that matter; the datasets only annotate amodal masks inside the frame and never for objects that leave the frame or disappear for a stretch, so the method's claimed robustness to occlusion is demonstrated only for objects that remain partially visible within the frame.

Editorial extensions

If this is right

  • On FISHBOWL with ResNet-50, A2VIS raises AP from 40.04 to 41.77 and IDF1 from 50.08 to 52.14 while cutting identity switches from 3480 to 3392 relative to GenVIS.
  • On SAIL VOS with ResNet-50, A2VIS improves AP to 23.12, HOTA to 30.04, and IDF1 to 25.94 relative to GenVIS, with fewer identity switches.
  • Under occlusion rates above 50%, A2VIS keeps an advantage over GenVIS-Amodal on FISHBOWL, with AP 33.14 versus 29.78.
  • Ablations show that both the visible spatiotemporal-prior mask and the amodal spatiotemporal-prior mask contribute to the gain; removing either lowers amodal VIS performance.
  • Adding SAMH improves even visible-only VIS metrics, with FISHBOWL AP rising from 39.94 to 41.77, suggesting amodal supervision helps ordinary tracking, not just amodal mask prediction.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the reported gains come from amodal masks being more temporally stable than visible masks, then the same mechanism should help any tracking-by-query method, not only prototype-based VIS; a direct test is to add SAMH to MOTR-style query trackers and compare identity-switch rates.
  • The paper leaves out-of-frame occlusion untested; a natural extension is to let amodal masks extend beyond the frame boundary using an external memory of object appearance, which would address the main real-world failure mode.
  • Because the authors report results on synthetic and game-derived datasets plus a zero-shot real-world probe, the strongest validation would come from a real-world dataset with dense amodal annotations; until such data exists, the practical benefit under realistic heavy occlusion remains an open question.
  • The ablation showing SAMH improves even visible-only metrics suggests amodal supervision acts as a regularizer; one could test whether weaker forms of full-shape supervision, such as amodal bounding-box expansion, give part of the same tracking gain at lower annotation cost.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper introduces A2VIS, a video instance segmentation framework that augments instance prototypes with amodal masks, using a Spatiotemporal-Prior Amodal Mask Head (SAMH) that combines short-range visible priors (VSPM) and long-range amodal priors (ASPM) via masked attention. The method is trained with both visible and amodal mask losses and evaluated on FISHBOWL and SAIL-VOS for VIS, amodal VIS, and MOT, plus a zero-shot OVIS evaluation. The authors report consistent improvements over published baselines and ablations showing the contribution of both spatiotemporal priors.

Significance. If the reported results are reliable, the paper demonstrates a useful inductive bias: incorporating amodal segmentation into instance prototypes improves occlusion robustness in video instance segmentation and tracking. The method is technically novel in combining spatiotemporal masked attention for amodal prediction with global instance prototypes, and the ablations support the role of the two priors. The paper also introduces a new set of amodal-VIS baselines. However, the evaluation scope is narrower than the abstract claims, and the lack of statistical rigor and validation-based hyperparameter selection weakens the strength of the conclusions; the absence of code is an additional reproducibility concern.

major comments (3)
  1. [§3.1, Limitation, Abstract] The amodal supervision is defined as frame-confined and empty until first visible appearance (§3.1), and the Limitation explicitly states that out-of-frame occlusion and disappear-and-reappear cases are not handled. Yet the Abstract and Introduction claim a 'reliable and comprehensive understanding' of objects that are 'partially or completely hidden', and Figure 1 promises 'complete trajectory and shape'. As a result, the experimental evidence in Tables 1-3 and 9 supports improvements for in-frame partial and full occlusion by other objects, but not for exit/re-entry or out-of-frame occlusion, which are dominant real-world tracking failure modes. Please either temper the claims to in-frame occlusion or provide experiments that cover complete disappearance and reappearance.
  2. [§4.4.1, Tables 1-8] All main results are reported as single runs, with no error bars, confidence intervals, or significance tests, making the phrase 'significant performance gap' in §4.4.1 unsupported. Moreover, the hyperparameters Nc, L, and the number of convolution layers are chosen based on ablations in Tables 5-7, yet there is no mention of a validation split; if the ablations were performed on the test set, the reported numbers are optimistically biased. Please report mean and variance over multiple seeds and clearly state the protocol for model selection.
  3. [§4.3, Table 2] The amodal baselines (Mask2Former-Amodal, VITA-Amodal, GenVIS-Amodal) are trained with amodal supervision alone, whereas A2VIS is trained with both visible and amodal losses (Eq. 4). This gives A2VIS additional supervision and makes the comparison in Table 2 not an apples-to-apples test of the proposed SAMH. To isolate the contribution of the amodal head, the authors should include multi-task baselines that are also trained with both visible and amodal masks, or otherwise match the supervision available to each method.
minor comments (5)
  1. [§3.5, Algorithm 1] Algorithm 1 uses ρ(Fk) for the Amodal Feature Extraction, while the text refers to Ω; please unify the notation.
  2. [§4.3] The description of the amodal baseline training is confusing: 'all the models are pretrained with frame-level FISHBOWL and SAIL VOS datasets on FISHBOWL dataset with amodal segmentation ground truth' seems to contain a typo and should be rewritten.
  3. [§4.3, Table 3] The MOT baseline list mentions TrackFormer and MOTR, but Table 3 reports MOTRv2; please add MOTRv2 to the baseline description for consistency.
  4. [§4.4.1, Figure 6 caption] The caption contains the typo 'quatitatively'; it should read 'qualitatively'.
  5. [§4.2] The implementation details state that a frame-level Mask2Former is pretrained and then video-level training is performed; please explain why this two-stage procedure is needed and how it interacts with the amodal supervision.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; external benchmarks and held-out tests ground the central claim, and the ASPM self-prior is an iterative refinement, not a fitted-input prediction.

full rationale

A2VIS's central derivation is not circular. The method is trained on FISHBOWL and SAIL-VOS with held-out test splits and compared against external published baselines (GenVIS, VITA, Mask2Former-VIS, DVIS, etc.), so the main performance claims do not reduce to the paper's own definitions. The only self-referential component is the Amodal Spatiotemporal-Prior Mask (ASPM) in Algorithm 1: the attention mask Tk is computed from the model's own predicted amodal segmentation Ak. This is a recurrent self-conditioning/iterative refinement mechanism, not a definitional equivalence; Table 4 shows empirically that ASPM and VSPM improve results, which is evidence it is a learned component rather than a renamed input. The paper also cites the authors' earlier AISFormer [37] and Shapeformer [36], but only for architectural choices (e.g., 'we follow [37, 7] to design Ω') and as a related-method baseline; these citations are not load-bearing for the central claim. The scope limitation in Section 5 is explicit: the method 'does not explicitly account for objects that are occluded by being partly or completely out of the frame or disappear in one frame and reappear in another.' This bounds the real-world generality of the claim but does not make the derivation circular, because the reported experiments and ablations operate under the same in-frame occlusion definition as the method's stated assumptions. Overall, no prediction in the paper reduces by construction to a fitted parameter or to the target quantity it claims to predict.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a small set of empirically chosen hyperparameters and on the validity of the synthetic amodal annotations. No new physical entities or external benchmark constants are introduced.

free parameters (3)
  • clip_length Nc = 3
    Selected based on ablation on FISHBOWL (§4.6.2, Table 5); highest scores at 3.
  • number of decoding layers L = 2
    Selected based on ablation (§4.6.3, Table 6) balancing AP and complexity.
  • number of conv layers in Amodal Feature Extraction = 4
    Selected based on ablation (§4.6.4, Table 7).
assumptions (4)
  • domain assumption Amodal ground truth in FISHBOWL and SAIL-VOS is correct and consistent.
    The model is trained and evaluated against these annotations; no independent verification is provided.
  • domain assumption Synthetic datasets (FISHBOWL, SAIL-VOS) represent real-world occlusion sufficiently for the claims.
    The paper acknowledges there is no real-world Amodal VIS dataset, so generalization is assumed.
  • ad hoc to paper The spatiotemporal-prior mask Tk computed from the union of visible and predicted amodal masks is a valid attention prior.
    Algorithm 1 gates attention to pixels inside visible or currently predicted amodal regions; this assumes the current prediction is close to the true shape.
  • domain assumption VITA and GenVIS are effective base models and their association mechanisms are reliable.
    A2VIS adopts VITA's prototype modelling (§3.3) and builds on GenVIS's framework; if these bases fail, A2VIS inherits the failure.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A2VIS: Amodal-Aware Approach to Video Instance Segmentation." pith.science (2026). https://pith.science/paper/GK2YZIZ3

@misc{pith2026241201147,
  author       = {Pith},
  title        = {Pith review of: A2VIS: Amodal-Aware Approach to Video Instance Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GK2YZIZ3}},
  note         = {Machine review of arXiv:2412.01147}
}
read the original abstract

Handling occlusion remains a significant challenge for video instance-level tasks like Multiple Object Tracking (MOT) and Video Instance Segmentation (VIS). In this paper, we propose a novel framework, Amodal-Aware Video Instance Segmentation (A2VIS), which incorporates amodal representations to achieve a reliable and comprehensive understanding of both visible and occluded parts of objects in a video. The key intuition is that awareness of amodal segmentation through spatiotemporal dimension enables a stable stream of object information. In scenarios where objects are partially or completely hidden from view, amodal segmentation offers more consistency and less dramatic changes along the temporal axis compared to visible segmentation. Hence, both amodal and visible information from all clips can be integrated into one global instance prototype. To effectively address the challenge of video amodal segmentation, we introduce the spatiotemporal-prior Amodal Mask Head, which leverages visible information intra clips while extracting amodal characteristics inter clips. Through extensive experiments and ablation studies, we show that A2VIS excels in both MOT and VIS tasks in identifying and tracking object instances with a keen understanding of their full shape.

Figures

Figures reproduced from arXiv: 2412.01147 by the authors.

Figure 1
Figure 1. Comparison between existing VIS and the proposed A2VIS. By integrating [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overall architecture of the proposed A2VIS. “IP” denotes instance prototypes [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 2
Figure 2. 3.3. Instance Prototype Modelling We adopt the object token association-based VITA as the clip-based in￾stance prototypes modelling Θ for its proven effectiveness and efficiency in modeling instance prototypes. This approach parses an input clip through object tokens without relying on a dense spatio-temporal backbone. It is advantageous for training on extended video sequences and facilitates estab￾lishing relation… view at source ↗
Figures from the paper (7 more)
Figure 3
Figure 3. Figure 3: Network design of Spatiotemporal-prior Amodal Mask Head (SAMH), which [PITH_FULL_IMAGE:figures/full_fig_p009_3.png]
Figure 4
Figure 4. Figure 4: Qualitative results of A2VIS on FISHBOWL dataset (first two rows) and SAIL [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparisons of A2VIS with GenVIS-Amodal. Videos are sourced [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison between our A2VIS and VITA and GenVIS on FISH [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: Qualitative comparison between our A2VIS and VITA-Amodal and GenVIS [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: A visual comparison between using cross-attention and spatiotemporal-prior [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: Qualitative outcomes of our A2VIS model, originally trained on SAILVOS, [PITH_FULL_IMAGE:figures/full_fig_p025_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 44 canonical work pages

  1. [1]

    Tarvis: A unified approach for target-based video segmentation, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Athar, A., Hermans, A., Luiten, J., Ramanan, D., Leibe, B., 2023. Tarvis: A unified approach for target-based video segmentation, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 18738–18748

    work page 2023

  2. [2]

    Athar, A., Mahadevan, S., Osep, A., Leal-Taix´ e, L., Leibe, B., 2020. Stem-seg: Spatio-temporal embeddings for instance segmentation in videos, in: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XI 16, Springer. pp. 158–177

    work page 2020

  3. [3]

    Memot: Multi-object tracking with memory, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Cai, J., Xu, M., Li, W., Xiong, Y., Xia, W., Tu, Z., Soatto, S., 2022. Memot: Multi-object tracking with memory, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 8090–8100

    work page 2022

  4. [4]

    Cao, J., Anwer, R.M., Cholakkal, H., Khan, F.S., Pang, Y., Shao, L.,

    work page

  5. [5]

    Observation-centric sort: Rethinking sort for robust multi-object track- ing, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Cao, J., Pang, J., Weng, X., Khirodkar, R., Kitani, K., 2023. Observation-centric sort: Rethinking sort for robust multi-object track- ing, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 9686–9696

    work page 2023

  6. [6]

    Mask2former for video instance segmentation

    Cheng, B., Choudhuri, A., Misra, I., Kirillov, A., Girdhar, R., Schwing, A.G., 2021. Mask2former for video instance segmentation. arXiv preprint arXiv:2112.10764

    arXiv 2021

  7. [7]

    Masked-attention mask transformer for universal image segmentation, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp

    Cheng, B., Misra, I., Schwing, A.G., Kirillov, A., Girdhar, R., 2022. Masked-attention mask transformer for universal image segmentation, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 1290–1299

    2022

  8. [8]

    Selective attention and the organization of visual information

    Duncan, J., 1984. Selective attention and the organization of visual information. Journal of experimental psychology: General 113, 501. 27

    work page 1984

Show all 52 references
  1. [9]

    Fan, K., Lei, J., Qian, X., Yu, M., Xiao, T., He, T., Zhang, Z., Fu, Y., 2023. Rethinking amodal video segmentation from learning super- vised signals with object-centric representation, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 1272– 1281

    2023

  2. [10]

    Follmann, P., K¨ onig, R., H¨ artinger, P., Klostermann, M., B¨ ottger, T.,

  3. [11]

    Coarse-to-fine amodal segmentation with shape prior, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Gao, J., Qian, X., Wang, Y., Xiao, T., He, T., Zhang, Z., Fu, Y., 2023. Coarse-to-fine amodal segmentation with shape prior, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 1262–1271

    2023

  4. [12]

    Mask r-cnn, in: Proceedings of the IEEE international conference on computer vision, pp

    He, K., Gkioxari, G., Doll´ ar, P., Girshick, R., 2017. Mask r-cnn, in: Proceedings of the IEEE international conference on computer vision, pp. 2961–2969

    2017

  5. [13]

    Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp

    He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778

    2016

  6. [14]

    A generalized framework for video instance segmentation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Heo, M., Hwang, S., Hyun, J., Kim, H., Oh, S.W., Lee, J.Y., Kim, S.J., 2023. A generalized framework for video instance segmentation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14623–14632

    2023

  7. [15]

    Vita: Video instance segmentation via object token association

    Heo, M., Hwang, S., Oh, S.W., Lee, J.Y., Kim, S.J., 2022. Vita: Video instance segmentation via object token association. Advances in Neural Information Processing Systems 35, 23109–23120

    2022

  8. [16]

    Hu, Y.T., Chen, H.S., Hui, K., Huang, J.B., Schwing, A.G., 2019. Sail- vos: Semantic amodal instance level video object segmentation-a syn- thetic dataset and baselines, in: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, pp. 3105–3115

    2019

  9. [17]

    Minvis: A minimal video instance segmentation framework without video-based training

    Huang, D.A., Yu, Z., Anandkumar, A., 2022. Minvis: A minimal video instance segmentation framework without video-based training. Ad- vances in Neural Information Processing Systems 35, 31265–31277. 28

    2022

  10. [18]

    Video instance seg- mentation using inter-frame communication transformers

    Hwang, S., Heo, M., Oh, S.W., Kim, S.J., 2021. Video instance seg- mentation using inter-frame communication transformers. Advances in Neural Information Processing Systems 34, 13352–13363

    2021

  11. [19]

    A theory of visual interpolation in object perception

    Kellman, P.J., Shipley, T.F., 1991. A theory of visual interpolation in object perception. Cognitive psychology 23, 141–221

    1991

  12. [20]

    Offline-to-online knowl- edge distillation for video instance segmentation, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp

    Kim, H., Lee, S., Kang, H., Im, S., 2024. Offline-to-online knowl- edge distillation for video instance segmentation, in: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp. 159–168

    2024

  13. [21]

    Amodal instance segmentation, in: European Conference on Computer Vision, Springer

    Li, K., Malik, J., 2016. Amodal instance segmentation, in: European Conference on Computer Vision, Springer. pp. 677–693

    2016

  14. [22]

    Microsoft coco: Common objects in context, in: European conference on computer vision, Springer

    Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Doll´ ar, P., Zitnick, C.L., 2014. Microsoft coco: Common objects in context, in: European conference on computer vision, Springer. pp. 740–755

    2014

  15. [23]

    Sg-net: Spatial granularity network for one-stage video instance segmentation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, pp

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

  16. [24]

    Blade: Box-level supervised amodal segmentation through directed expansion, in: Proceedings of the AAAI Conference on Artificial Intelligence, pp

    Liu, Z., Li, Z., Jiang, T., 2024. Blade: Box-level supervised amodal segmentation through directed expansion, in: Proceedings of the AAAI Conference on Artificial Intelligence, pp. 3846–3854

    2024

  17. [25]

    Swin transformer: Hierarchical vision transformer using shifted windows, in: Proceedings of the IEEE/CVF international conference on computer vision, pp

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

  18. [26]

    Hota: A higher order metric for evaluating multi-object tracking

    Luiten, J., Osep, A., Dendorfer, P., Torr, P., Geiger, A., Leal-Taix´ e, L., Leibe, B., 2021. Hota: A higher order metric for evaluating multi-object tracking. International journal of computer vision 129, 548–578

    2021

  19. [27]

    Trackformer: Multi-object tracking with transformers, in: Proceedings 29 of the IEEE/CVF conference on computer vision and pattern recogni- tion, pp

    Meinhardt, T., Kirillov, A., Leal-Taixe, L., Feichtenhofer, C., 2022. Trackformer: Multi-object tracking with transformers, in: Proceedings 29 of the IEEE/CVF conference on computer vision and pattern recogni- tion, pp. 8844–8854

    2022

  20. [28]

    Video object segmentation using space-time memory networks, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Oh, S.W., Lee, J.Y., Xu, N., Kim, S.J., 2019. Video object segmentation using space-time memory networks, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 9226–9235

    2019

  21. [29]

    pix2gestalt: Amodal segmentation by synthesiz- ing wholes, in: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE Computer Society

    Ozguroglu, E., Liu, R., Sur ´ ıs, D., Chen, D., Dave, A., Tokmakov, P., Vondrick, C., 2024. pix2gestalt: Amodal segmentation by synthesiz- ing wholes, in: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE Computer Society. pp. 3931–3940

    2024

  22. [30]

    Occluded video instance segmentation: A benchmark

    Qi, J., Gao, Y., Hu, Y., Wang, X., Liu, X., Bai, X., Belongie, S., Yuille, A., Torr, P.H., Bai, S., 2022. Occluded video instance segmentation: A benchmark. International Journal of Computer Vision 130, 2022–2039

    2022

  23. [31]

    Coarse-to- fine video instance segmentation with factorized conditional appearance flows

    Qin, Z., Lu, X., Nie, X., Liu, D., Yin, Y., Wang, W., 2023a. Coarse-to- fine video instance segmentation with factorized conditional appearance flows. IEEE/CAA Journal of Automatica Sinica 10, 1192–1208

  24. [32]

    Motiontrack: Learning robust short-term and long-term motions for multi-object tracking, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Qin, Z., Zhou, S., Wang, L., Duan, J., Hua, G., Tang, W., 2023b. Motiontrack: Learning robust short-term and long-term motions for multi-object tracking, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 17939–17948

  25. [33]

    Perfor- mance measures and a data set for multi-target, multi-camera tracking, in: European conference on computer vision, Springer

    Ristani, E., Solera, F., Zou, R., Cucchiara, R., Tomasi, C., 2016. Perfor- mance measures and a data set for multi-target, multi-camera tracking, in: European conference on computer vision, Springer. pp. 17–35

    2016

  26. [34]

    Dancetrack: Multi-object tracking in uniform appearance and diverse motion, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Sun, P., Cao, J., Jiang, Y., Yuan, Z., Bai, S., Kitani, K., Luo, P., 2022. Dancetrack: Multi-object tracking in uniform appearance and diverse motion, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 20993–21002

    2022

  27. [35]

    Unsuper- vised object learning via common fate

    Tangemann, M., Schneider, S., Von K¨ ugelgen, J., Locatello, F., Gehler, P., Brox, T., K¨ ummerer, M., Bethge, M., Sch¨ olkopf, B., 2021. Unsuper- vised object learning via common fate. arXiv preprint arXiv:2110.06562 . 30

    2021 arXiv

  28. [36]

    Shapeformer: Shape prior visible-to-amodal transformer-based amodal instance segmentation

    Tran, M., Bounsavy, W., Vo, K., Nguyen, A., Nguyen, T., Le, N., 2024. Shapeformer: Shape prior visible-to-amodal transformer-based amodal instance segmentation. arXiv preprint arXiv:2403.11376

    2024 arXiv

  29. [37]

    Aisformer: Amodal instance segmentation with transformer

    Tran, M., Vo, K., Yamazaki, K., Fernandes, A., Kidd, M., Le, N., 2022. Aisformer: Amodal instance segmentation with transformer. arXiv preprint arXiv:2210.06323

    2022 arXiv

  30. [38]

    Ov-vis: Open-vocabulary video instance segmentation

    Wang, H., Yan, C., Chen, K., Jiang, X., Tang, X., Hu, Y., Kang, G., Xie, W., Gavves, E., 2024. Ov-vis: Open-vocabulary video instance segmentation. International Journal of Computer Vision 132, 5048– 5065

    2024

  31. [39]

    Seqformer: Se- quential transformer for video instance segmentation, in: European Con- ference on Computer Vision, Springer

    Wu, J., Jiang, Y., Bai, S., Zhang, W., Bai, X., 2022a. Seqformer: Se- quential transformer for video instance segmentation, in: European Con- ference on Computer Vision, Springer. pp. 553–569

  32. [40]

    In de- fense of online models for video instance segmentation, in: European Conference on Computer Vision, Springer

    Wu, J., Liu, Q., Jiang, Y., Bai, S., Yuille, A., Bai, X., 2022b. In de- fense of online models for video instance segmentation, in: European Conference on Computer Vision, Springer. pp. 588–605

  33. [41]

    Amodal seg- mentation based on visible region segmentation and shape prior

    Xiao, Y., Xu, Y., Zhong, Z., Luo, W., Li, J., Gao, S., 2020. Amodal seg- mentation based on visible region segmentation and shape prior. arXiv preprint arXiv:2012.05598

    2020 arXiv

  34. [42]

    Video instance segmentation, in: Pro- ceedings of the IEEE/CVF International Conference on Computer Vi- sion, pp

    Yang, L., Fan, Y., Xu, N., 2019. Video instance segmentation, in: Pro- ceedings of the IEEE/CVF International Conference on Computer Vi- sion, pp. 5188–5197

    2019

  35. [43]

    Self-supervised amodal video object segmentation

    Yao, J., Hong, Y., Wang, C., Xiao, T., He, T., Locatello, F., Wipf, D.P., Fu, Y., Zhang, Z., 2022. Self-supervised amodal video object segmentation. Advances in Neural Information Processing Systems 35, 6278–6291

    2022

  36. [44]

    Motr: End-to-end multiple-object tracking with transformer, in: Euro- pean Conference on Computer Vision, Springer

    Zeng, F., Dong, B., Zhang, Y., Wang, T., Zhang, X., Wei, Y., 2022. Motr: End-to-end multiple-object tracking with transformer, in: Euro- pean Conference on Computer Vision, Springer. pp. 659–675

    2022

  37. [45]

    Amodal ground truth and completion in the wild, in: Proceedings of the IEEE/CVF 31 Conference on Computer Vision and Pattern Recognition, pp

    Zhan, G., Zheng, C., Xie, W., Zisserman, A., 2024. Amodal ground truth and completion in the wild, in: Proceedings of the IEEE/CVF 31 Conference on Computer Vision and Pattern Recognition, pp. 28003– 28013

    2024

  38. [46]

    Dvis: Decoupled video instance segmentation framework

    Zhang, T., Tian, X., Wu, Y., Ji, S., Wang, X., Zhang, Y., Wan, P., 2023a. Dvis: Decoupled video instance segmentation framework. arXiv preprint arXiv:2306.03413

    arXiv

  39. [47]

    Bytetrack: Multi-object tracking by associat- ing every detection box, in: European Conference on Computer Vision, Springer

    Zhang, Y., Sun, P., Jiang, Y., Yu, D., Weng, F., Yuan, Z., Luo, P., Liu, W., Wang, X., 2022. Bytetrack: Multi-object tracking by associat- ing every detection box, in: European Conference on Computer Vision, Springer. pp. 1–21

    2022

  40. [48]

    Fairmot: On the fairness of detection and re-identification in multiple object tracking

    Zhang, Y., Wang, C., Wang, X., Zeng, W., Liu, W., 2021. Fairmot: On the fairness of detection and re-identification in multiple object tracking. International Journal of Computer Vision 129, 3069–3087

    2021

  41. [49]

    Motrv2: Bootstrapping end-to- end multi-object tracking by pretrained object detectors, in: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Zhang, Y., Wang, T., Zhang, X., 2023b. Motrv2: Bootstrapping end-to- end multi-object tracking by pretrained object detectors, in: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 22056–22065

  42. [50]

    Deformable detr: Deformable transformers for end-to-end object detection

    Zhu, X., Su, W., Lu, L., Li, B., Wang, X., Dai, J., 2020. Deformable detr: Deformable transformers for end-to-end object detection. arXiv preprint arXiv:2010.04159 . 32

    2020 arXiv

  43. [2019]

    Learning to see the invisible: End-to-end trainable amodal in- stance segmentation, in: 2019 IEEE Winter Conference on Applications of Computer Vision (W ACV), IEEE. pp. 1328–1336

    2019

  44. [2020]

    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, Springer. pp. 1–18

    2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.