pith. sign in

arxiv: 2606.12826 · v1 · pith:HODHTGVBnew · submitted 2026-06-11 · 💻 cs.CV · cs.AI

DIMOS: Disentangling Instance-level Moving Object Segmentation

Hongxiang Huang , Hongwei Ren , Xiaopeng Lin , Yulong Huang , Zeke Xie , Bojun Cheng This is my paper

Pith reviewed 2026-06-27 07:50 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords moving instance segmentationevent cameramultimodal fusionfeature disentanglementcross-modal alignmentsmall object detection
0
0 comments X

The pith

A dual-disentangling framework separates appearance and motion in image and event modalities to improve segmentation of small moving instances.

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

The paper introduces DIMOS to address limitations in multimodal moving instance segmentation by fusing standard camera images with event camera data. It proposes a dual-disentangling feature extraction step that isolates appearance attributes from motion cues inside each modality separately. A subsequent multi-granularity cross-modal alignment step then matches the resulting features for fusion. This produces denser representations that handle sparse event data and small instances more effectively. Experiments show gains over prior methods particularly in fast-motion and low-light conditions.

Core claim

Separating appearance and motion information within both image and event modalities, followed by distributionally and semantically consistent cross-modal alignment at multiple granularities, yields fused features that enable state-of-the-art moving instance segmentation performance, especially for small instances under fast motion and low-light conditions.

What carries the argument

The dual-disentangling feature extraction framework that isolates appearance from motion in each modality, combined with multi-granularity cross-modal alignment for fusion.

If this is right

  • Small moving objects become detectable in sparse event streams when motion cues are isolated from appearance.
  • Cross-modal fusion gains reliability once features are aligned both statistically and semantically at multiple scales.
  • Performance holds under fast motion and low light because motion information is no longer diluted by appearance entanglement.
  • The approach extends directly to traffic surveillance and autonomous driving where small distant objects matter.

Where Pith is reading between the lines

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

  • The same separation step could be tested on other event-plus-image tasks such as optical flow or depth estimation.
  • If the disentangled features prove more compact, downstream models might run at lower compute cost without accuracy loss.
  • Extending the alignment to additional modalities like lidar would check whether the framework generalizes beyond two sensors.

Load-bearing premise

Separating appearance and motion inside each modality produces denser features that remain complete enough for effective cross-modal fusion.

What would settle it

A controlled comparison in which entangled event and image features achieve equal or higher segmentation accuracy on small instances than the disentangled versions.

Figures

Figures reproduced from arXiv: 2606.12826 by Bojun Cheng, Hongwei Ren, Hongxiang Huang, Xiaopeng Lin, Yulong Huang, Zeke Xie.

Figure 1
Figure 1. Figure 1: Feature Extraction Comparison. (a) Our method extracts [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the proposed DIMOS framework. The pipeline consists of four major components: (1) Dual-Disentangling Mechanism with appearance and motion encoders for each modality. (2) Multi-Granularity Cross-Modal Alignment & Fusion that enforces consistency at both distributional and semantical levels. (3) Cross-Type Interaction for joint reasoning between appearance and motion cues via cross attention. and… view at source ↗
Figure 3
Figure 3. Figure 3: Visual comparisons show consecutive frames sampled from a video sequence of MouseSIS [ [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Intra-modal contrastive learning for strengthening [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Effect of intra-modal contrastive learning on [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Consecutive frames sampled from a video sequence of SEVD-Fixed are arranged from top to bottom. Red boxes highlight [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
read the original abstract

Moving instance segmentation (MIS) attracts increasing attention due to its broad applications in traffic surveillance, autonomous driving, and animal tracking. Event cameras record asynchronous brightness changes, providing high temporal resolution and dynamic range, which makes them highly sensitive to motion information. By fusing event and image features, motion cues from events can complement spatial details from images, enhancing the performance of MIS. However, current multimodal MIS methods still struggle to segment small moving instances, as event cameras often yield sparse features under limited resolution. Moreover, event features entangle appearance attributes with motion cues, which further restricts effective cross-modal fusion. To address these challenges, we first propose a dual-disentangling feature extraction framework that separates and extracts appearance and motion information within both image and event modalities, thereby improving feature density. Subsequently, a multi-granularity cross-modal alignment is introduced to align distributionally and semantically consistent features across modalities, enabling more effective fusion with rich spatial and temporal details. The experiment results demonstrate that our method achieves state-of-the-art performance in multimodal MIS, especially for small instances under challenging conditions such as fast motion and low-light settings.

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 / 0 minor

Summary. The paper proposes DIMOS, a dual-disentangling feature extraction framework that separates appearance and motion cues within both image and event modalities to improve feature density for moving instance segmentation (MIS), followed by a multi-granularity cross-modal alignment module to enable effective fusion. It claims state-of-the-art performance on multimodal MIS, with particular gains for small instances under fast motion and low-light conditions.

Significance. If the performance claims hold after proper validation, the dual-disentangling strategy could meaningfully advance multimodal MIS by addressing sparsity and entanglement issues in event data, offering denser features for small-object cases that are critical in applications such as autonomous driving and surveillance.

major comments (2)
  1. [Abstract] Abstract: The central claim that 'the experiment results demonstrate that our method achieves state-of-the-art performance' is unsupported; the manuscript supplies no experimental protocol, metrics, baselines, datasets, ablation studies, or quantitative results, making the SOTA assertion unverifiable and load-bearing for the paper's contribution.
  2. [Abstract] Abstract (paragraph on dual-disentangling framework): No explicit argument, mathematical definition, or empirical verification is provided that the appearance/motion separation operators preserve low-amplitude motion signals and fine appearance gradients required for small-instance segmentation; if the disentangling implicitly thresholds or projects away such cues, downstream fusion cannot recover them, directly undermining the headline gains under challenging conditions.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the abstract. We address the two major comments point by point below, proposing revisions to strengthen verifiability and clarity where the current text falls short.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central claim that 'the experiment results demonstrate that our method achieves state-of-the-art performance' is unsupported; the manuscript supplies no experimental protocol, metrics, baselines, datasets, ablation studies, or quantitative results, making the SOTA assertion unverifiable and load-bearing for the paper's contribution.

    Authors: The provided manuscript text consists solely of the abstract, which indeed contains no experimental protocol, metrics, baselines, datasets, ablation studies, or quantitative results. The SOTA claim therefore cannot be verified from the given text. We will revise the abstract to remove the unsupported claim or qualify it pending addition of a concise results summary (e.g., key mIoU/AP numbers and dataset names) in a revised version. revision: yes

  2. Referee: [Abstract] Abstract (paragraph on dual-disentangling framework): No explicit argument, mathematical definition, or empirical verification is provided that the appearance/motion separation operators preserve low-amplitude motion signals and fine appearance gradients required for small-instance segmentation; if the disentangling implicitly thresholds or projects away such cues, downstream fusion cannot recover them, directly undermining the headline gains under challenging conditions.

    Authors: The provided manuscript text supplies no mathematical definitions, arguments, or empirical verification that the separation operators preserve low-amplitude signals and fine gradients. This is a substantive gap in the abstract. We will revise the abstract to include a brief description of the operators' design intent and will add an ablation study in the full revision to empirically demonstrate preservation of these cues for small instances. revision: yes

Circularity Check

0 steps flagged

No circularity detected in derivation chain

full rationale

The provided abstract and method outline describe an architectural proposal (dual-disentangling feature extraction plus multi-granularity alignment) whose core steps are presented as design choices rather than derived quantities. No equations, fitted parameters, or predictions are shown that reduce by construction to their own inputs. No self-citation chains or uniqueness theorems are invoked to justify the framework. The SOTA performance claim rests on experimental results, which remain externally falsifiable and independent of the framework definition. This is the normal case of a self-contained empirical method paper.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract supplies no explicit free parameters, mathematical axioms, or newly postulated entities; all technical content remains at the level of named modules.

pith-pipeline@v0.9.1-grok · 5735 in / 1039 out tokens · 21524 ms · 2026-06-27T07:50:35.555264+00:00 · methodology

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

58 extracted references · 2 linked inside Pith

  1. [1]

    Sevd: Synthetic event-based vision dataset for ego and fixed traffic perception.arXiv preprint arXiv:2404.10540, 2024

    Manideep Reddy Aliminati, Bharatesh Chakravarthi, Aayush Atul Verma, Arpitsinh Vaghela, Hua Wei, Xuesong Zhou, and Yezhou Yang. Sevd: Synthetic event-based vision dataset for ego and fixed traffic perception.arXiv preprint arXiv:2404.10540, 2024. 6

    arXiv 2024

  2. [2]

    Foreground segmen- tation using a triplet convolutional neural network for mul- tiscale feature encoding.arXiv e-prints, pages arXiv–1801,

    Long Ang Lim and Hacer Yalim Keles. Foreground segmen- tation using a triplet convolutional neural network for mul- tiscale feature encoding.arXiv e-prints, pages arXiv–1801,

  3. [3]

    Per- pixel classification is not all you need for semantic segmen- tation.Advances in neural information processing systems, 34:17864–17875, 2021

    Bowen Cheng, Alex Schwing, and Alexander Kirillov. Per- pixel classification is not all you need for semantic segmen- tation.Advances in neural information processing systems, 34:17864–17875, 2021. 4, 3

    2021

  4. [4]

    Masked-attention mask transformer for universal image segmentation

    Bowen Cheng, Ishan Misra, Alexander G Schwing, Alexan- der Kirillov, and Rohit Girdhar. Masked-attention mask transformer for universal image segmentation. InProceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1290–1299, 2022. 3

    2022

  5. [5]

    Xmem: Long- term video object segmentation with an atkinson-shiffrin memory model

    Ho Kei Cheng and Alexander G Schwing. Xmem: Long- term video object segmentation with an atkinson-shiffrin memory model. InEuropean conference on computer vision, pages 640–658. Springer, 2022. 1

    2022

  6. [6]

    Disentangling writer and character styles for handwriting generation

    Gang Dai, Yifan Zhang, Qingfeng Wang, Qing Du, Zhuliang Yu, Zhuoman Liu, and Shuangping Huang. Disentangling writer and character styles for handwriting generation. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 5977–5986, 2023. 2

    2023

  7. [7]

    One-dm: One-shot diffusion mimicker for handwritten text generation

    Gang Dai, Yifan Zhang, Quhui Ke, Qiangya Guo, and Shuangping Huang. One-dm: One-shot diffusion mimicker for handwritten text generation. InEuropean Conference on Computer Vision, pages 410–427. Springer, 2024. 3

    2024

  8. [8]

    Vg-sam: Visual in-context guided sam for universal medical image segmentation.Fractal and Frac- tional, 9(11):722, 2025

    Gang Dai, Qingfeng Wang, Yutao Qin, Gang Wei, and Shuangping Huang. Vg-sam: Visual in-context guided sam for universal medical image segmentation.Fractal and Frac- tional, 9(11):722, 2025. 3

    2025

  9. [9]

    Beyond isolated words: Diffusion brush for handwritten text-line generation

    Gang Dai, Yifan Zhang, Yutao Qin, Qiangya Guo, Shuang- ping Huang, and Shuicheng Yan. Beyond isolated words: Diffusion brush for handwritten text-line generation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 19054–19064, 2025. 2

    2025

  10. [10]

    Flownet: Learning optical flow with convolutional networks

    Alexey Dosovitskiy, Philipp Fischer, Eddy Ilg, Philip Hausser, Caner Hazirbas, Vladimir Golkov, Patrick Van Der Smagt, Daniel Cremers, and Thomas Brox. Flownet: Learning optical flow with convolutional networks. InPro- ceedings of the IEEE international conference on computer vision, pages 2758–2766, 2015. 2

    2015

  11. [11]

    Spikeram: A 48.1 pw/synapse/bit event-driven spiking compute-near/in- memory processor with neuromorphic sensor enabling life- long on-chip learning

    Haotian Fu, Yue Zhou, Zhuo Zhang, Hongzhao Zheng, Renxu Yang, Yulong Huang, Dezhen Yang, Yannan Xing, Tugba Demirci, Ning Qiao, et al. Spikeram: A 48.1 pw/synapse/bit event-driven spiking compute-near/in- memory processor with neuromorphic sensor enabling life- long on-chip learning. In2026 IEEE International Solid- State Circuits Conference (ISSCC), page...

  12. [12]

    A unifying contrast maximization framework for event cam- eras, with applications to motion, depth, and optical flow estimation

    Guillermo Gallego, Henri Rebecq, and Davide Scaramuzza. A unifying contrast maximization framework for event cam- eras, with applications to motion, depth, and optical flow estimation. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 3867–3876,

  13. [13]

    Event-based vision: A survey.IEEE transactions on pattern analysis and machine intelligence, 44(1):154–180, 2020

    Guillermo Gallego, Tobi Delbr ¨uck, Garrick Orchard, Chiara Bartolozzi, Brian Taba, Andrea Censi, Stefan Leutenegger, Andrew J Davison, J ¨org Conradt, Kostas Daniilidis, et al. Event-based vision: A survey.IEEE transactions on pattern analysis and machine intelligence, 44(1):154–180, 2020. 2

    2020

  14. [14]

    From motion blur to motion flow: A deep learning so- lution for removing heterogeneous motion blur

    Dong Gong, Jie Yang, Lingqiao Liu, Yanning Zhang, Ian Reid, Chunhua Shen, Anton Van Den Hengel, and Qinfeng Shi. From motion blur to motion flow: A deep learning so- lution for removing heterogeneous motion blur. InProceed- ings of the IEEE conference on computer vision and pattern recognition, pages 2319–2328, 2017. 2

    2017

  15. [15]

    Mousesis: A frames-and- events dataset for space-time instance segmentation of mice

    Friedhelm Hamann, Hanxiong Li, Paul Mieske, Lars Lewe- johann, and Guillermo Gallego. Mousesis: A frames-and- events dataset for space-time instance segmentation of mice. InEuropean Conference on Computer Vision, pages 156–

  16. [16]

    1, 3, 6, 7, 8

    Springer, 2024. 1, 3, 6, 7, 8

    2024

  17. [17]

    Sis-challenge: Event-based spatio-temporal instance segmentation chal- lenge at the cvpr 2025 event-based vision workshop

    Friedhelm Hamann, Emil Mededovic, Fabian G ¨ulhan, Yuli Wu, Johannes Stegmaier, Jing He, Yiqing Wang, Kexin Zhang, Lingling Li, Licheng Jiao, et al. Sis-challenge: Event-based spatio-temporal instance segmentation chal- lenge at the cvpr 2025 event-based vision workshop. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 467...

    2025

  18. [18]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 8

    2016

  19. [19]

    Exploring temporal dynamics in event- based eye tracker

    Hongxiang Huang, Xiaopeng Lin, Hongwei Ren, Yue Zhou, and Bojun Cheng. Exploring temporal dynamics in event- based eye tracker. InProceedings of the Computer Vi- sion and Pattern Recognition Conference, pages 5145–5154,

  20. [20]

    Clif: Complementary leaky integrate-and-fire neuron for spiking neural networks

    Yulong Huang, Xiaopeng Lin, Hongwei Ren, Haotian Fu, Yue Zhou, Zunchang Liu, Biao Pan, and Bojun Cheng. Clif: Complementary leaky integrate-and-fire neuron for spiking neural networks. InInternational Conference on Machine Learning, pages 19949–19972. PMLR, 2024. 1

    2024

  21. [21]

    Adam: A method for stochastic opti- mization.arXiv preprint arXiv:1412.6980, 2014

    Diederik P Kingma. Adam: A method for stochastic opti- mization.arXiv preprint arXiv:1412.6980, 2014. 6

    Pith/arXiv arXiv 2014

  22. [22]

    Segment any- thing

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C Berg, Wan-Yen Lo, et al. Segment any- thing. InProceedings of the IEEE/CVF international confer- ence on computer vision, pages 4015–4026, 2023. 3

    2023

  23. [23]

    Dam-vsr: Disentanglement of appear- ance and motion for video super-resolution

    Zhe Kong, Le Li, Yong Zhang, Feng Gao, Shaoshu Yang, Tao Wang, Kaihao Zhang, Zhuoliang Kang, Xiaoming Wei, Guanying Chen, et al. Dam-vsr: Disentanglement of appear- ance and motion for video super-resolution. InProceedings of the Special Interest Group on Computer Graphics and In- teractive Techniques Conference Conference Papers, pages 1–11, 2025. 2

    2025

  24. [24]

    Event-assisted low-light video object segmentation

    Hebei Li, Jin Wang, Jiahui Yuan, Yue Li, Wenming Weng, Yansong Peng, Yueyi Zhang, Zhiwei Xiong, and Xiaoyan Sun. Event-assisted low-light video object segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 3250–3259, 2024. 1, 3

    2024

  25. [25]

    Efficient event-based se- mantic segmentation via exploiting frame-event fusion: A hybrid neural network approach

    Hebei Li, Yansong Peng, Jiahui Yuan, Peixi Wu, Jin Wang, Yueyi Zhang, and Xiaoyan Sun. Efficient event-based se- mantic segmentation via exploiting frame-event fusion: A hybrid neural network approach. InProceedings of the AAAI Conference on Artificial Intelligence, pages 18296–18304,

  26. [26]

    A 128×128 120 dB 15µslatency asynchronous temporal con- trast vision sensor.IEEE journal of solid-state circuits, 43 (2):566–576, 2008

    Patrick Lichtsteiner, Christoph Posch, and Tobi Delbruck. A 128×128 120 dB 15µslatency asynchronous temporal con- trast vision sensor.IEEE journal of solid-state circuits, 43 (2):566–576, 2008. 1

    2008

  27. [27]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In European conference on computer vision, pages 740–755. Springer, 2014. 6

    2014

  28. [28]

    Clearsight: Human vision-inspired solutions for event-based motion deblurring

    Xiaopeng Lin, Yulong Huang, Hongwei Ren, Zunchang Liu, Hongxiang Huang, Yue Zhou, Haotian Fu, and Bojun Cheng. Clearsight: Human vision-inspired solutions for event-based motion deblurring. InProceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 7462–7471,

  29. [29]

    Event- based motion deblurring via multi-temporal granularity fu- sion.IEEE Transactions on Circuits and Systems for Video Technology, 2026

    Xiaopeng Lin, Hongwei Ren, Yulong Huang, Zunchang Liu, Yue Zhou, Haotian Fu, Biao Pan, and Bojun Cheng. Event- based motion deblurring via multi-temporal granularity fu- sion.IEEE Transactions on Circuits and Systems for Video Technology, 2026. 1

    2026

  30. [30]

    Ddflow: Learning optical flow with unlabeled data distilla- tion

    Pengpeng Liu, Irwin King, Michael R Lyu, and Jia Xu. Ddflow: Learning optical flow with unlabeled data distilla- tion. InProceedings of the AAAI conference on artificial intelligence, pages 8770–8777, 2019. 5

    2019

  31. [31]

    Ev-imo: Motion seg- mentation dataset and learning pipeline for event cameras

    Anton Mitrokhin, Chengxi Ye, Cornelia Ferm ¨uller, Yian- nis Aloimonos, and Tobi Delbruck. Ev-imo: Motion seg- mentation dataset and learning pipeline for event cameras. In2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 6105–6112. IEEE, 2019. 3, 6

    2019

  32. [32]

    Learning visual motion segmentation using event surfaces

    Anton Mitrokhin, Zhiyuan Hua, Cornelia Fermuller, and Yiannis Aloimonos. Learning visual motion segmentation using event surfaces. InProceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 14414–14423, 2020. 2, 3

    2020

  33. [33]

    Repre- sentation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748, 2018

    Aaron van den Oord, Yazhe Li, and Oriol Vinyals. Repre- sentation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748, 2018. 3, 4

    Pith/arXiv arXiv 2018

  34. [34]

    Disentangle domain features for cross-modality cardiac im- age segmentation.Medical Image Analysis, 71:102078,

    Chenhao Pei, Fuping Wu, Liqin Huang, and Xiahai Zhuang. Disentangle domain features for cross-modality cardiac im- age segmentation.Medical Image Analysis, 71:102078,

  35. [35]

    Optical flow estima- tion using a spatial pyramid network

    Anurag Ranjan and Michael J Black. Optical flow estima- tion using a spatial pyramid network. InProceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 4161–4170, 2017. 2

    2017

  36. [36]

    You only look once: Unified, real-time object de- tection

    Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. You only look once: Unified, real-time object de- tection. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 779–788, 2016. 3

    2016

  37. [37]

    E2b: A single modality point-based tracker with event cameras

    Hongwei Ren, Zhuo Li, Aiersi Tuerhong, Haobo Liu, Fei Liang, Yongxiang Feng, Wenhui Wang, Yaoyuan Wang, Ziyang Zhang, Weihua He, et al. E2b: A single modality point-based tracker with event cameras. In2025 IEEE In- ternational Conference on Robotics and Automation (ICRA), pages 6461–6468. IEEE, 2025. 1

    2025

  38. [38]

    Rethinking efficient and effective point- based networks for event camera classification and regres- sion.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Hongwei Ren, Yue Zhou, Jiadong Zhu, Xiaopeng Lin, Hao- tian Fu, Yulong Huang, Yuetong Fang, Fei Ma, Hao Yu, and Bojun Cheng. Rethinking efficient and effective point- based networks for event camera classification and regres- sion.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 1

    2025

  39. [39]

    Mobilenetv2: Inverted residuals and linear bottlenecks

    Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zh- moginov, and Liang-Chieh Chen. Mobilenetv2: Inverted residuals and linear bottlenecks. InProceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 4510–4520, 2018. 8

    2018

  40. [40]

    Adversarial discriminative domain adaptation

    Eric Tzeng, Judy Hoffman, Kate Saenko, and Trevor Darrell. Adversarial discriminative domain adaptation. InProceed- ings of the IEEE conference on computer vision and pattern recognition, pages 7167–7176, 2017. 5

    2017

  41. [41]

    Feelvos: Fast end-to-end embedding learning for video object seg- mentation

    Paul V oigtlaender, Yuning Chai, Florian Schroff, Hartwig Adam, Bastian Leibe, and Liang-Chieh Chen. Feelvos: Fast end-to-end embedding learning for video object seg- mentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 9481–9490,

  42. [42]

    Instance-level moving object segmentation from a sin- gle image with events.International Journal of Computer Vision, pages 1–22, 2025

    Zhexiong Wan, Bin Fan, Le Hui, Yuchao Dai, and Gim Hee Lee. Instance-level moving object segmentation from a sin- gle image with events.International Journal of Computer Vision, pages 1–22, 2025. 2, 3, 4, 5, 6, 7

    2025

  43. [43]

    Disentangling light fields for super-resolution and disparity estimation

    Yingqian Wang, Longguang Wang, Gaochang Wu, Jungang Yang, Wei An, Jingyi Yu, and Yulan Guo. Disentangling light fields for super-resolution and disparity estimation. IEEE Transactions on Pattern Analysis and Machine Intel- ligence, 45(1):425–443, 2022. 2

    2022

  44. [44]

    Un- evimo: Unsupervised event-based independent motion seg- mentation

    Ziyun Wang, Jinyuan Guo, and Kostas Daniilidis. Un- evimo: Unsupervised event-based independent motion seg- mentation. InEuropean Conference on Computer Vision, pages 228–245. Springer, 2024. 3

    2024

  45. [45]

    Disentangle then parse: Night-time se- mantic segmentation with illumination disentanglement

    Zhixiang Wei, Lin Chen, Tao Tu, Pengyang Ling, Huaian Chen, and Yi Jin. Disentangle then parse: Night-time se- mantic segmentation with illumination disentanglement. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 21593–21603, 2023. 2

    2023

  46. [46]

    Seqformer: Sequential transformer for video instance segmentation

    Junfeng Wu, Yi Jiang, Song Bai, Wenqing Zhang, and Xiang Bai. Seqformer: Sequential transformer for video instance segmentation. InEuropean Conference on Computer Vision, pages 553–569. Springer, 2022. 3

    2022

  47. [47]

    In defense of online models for video instance segmentation

    Junfeng Wu, Qihao Liu, Yi Jiang, Song Bai, Alan Yuille, and Xiang Bai. In defense of online models for video instance segmentation. InEuropean Conference on Computer Vision, pages 588–605. Springer, 2022. 1, 3, 6, 7

    2022

  48. [48]

    Eisnet: A multi-modal fusion network for semantic segmen- tation with events and images.IEEE Transactions on Multi- media, 26:8639–8650, 2024

    Bochen Xie, Yongjian Deng, Zhanpeng Shao, and Youfu Li. Eisnet: A multi-modal fusion network for semantic segmen- tation with events and images.IEEE Transactions on Multi- media, 26:8639–8650, 2024. 2

    2024

  49. [49]

    Collaborative video object segmentation by foreground-background inte- gration

    Zongxin Yang, Yunchao Wei, and Yi Yang. Collaborative video object segmentation by foreground-background inte- gration. InEuropean Conference on Computer Vision, pages 332–348. Springer, 2020. 2

    2020

  50. [50]

    Collabora- tive video object segmentation by multi-scale foreground- background integration.IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 44(9):4701–4712, 2021

    Zongxin Yang, Yunchao Wei, and Yi Yang. Collabora- tive video object segmentation by multi-scale foreground- background integration.IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 44(9):4701–4712, 2021. 2

    2021

  51. [51]

    Temporal-wise at- tention spiking neural networks for event streams classifica- tion

    Man Yao, Huanhuan Gao, Guangshe Zhao, Dingheng Wang, Yihan Lin, Zhaoxu Yang, and Guoqi Li. Temporal-wise at- tention spiking neural networks for event streams classifica- tion. InProceedings of the IEEE/CVF international confer- ence on computer vision, pages 10221–10230, 2021. 1

    2021

  52. [52]

    Eventpsr: Surface normal and reflectance estimation from photomet- ric stereo using an event camera

    Bohan Yu, Jin Han, Boxin Shi, and Imari Sato. Eventpsr: Surface normal and reflectance estimation from photomet- ric stereo using an event camera. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 11427–11436, 2025. 2

    2025

  53. [53]

    Isomer: Isomerous transformer for zero-shot video object segmenta- tion

    Yichen Yuan, Yifan Wang, Lijun Wang, Xiaoqi Zhao, Huchuan Lu, Yu Wang, Weibo Su, and Lei Zhang. Isomer: Isomerous transformer for zero-shot video object segmenta- tion. InProceedings of the IEEE/CVF international confer- ence on computer vision, pages 966–976, 2023. 2

    2023

  54. [54]

    Radar instance transformer: Reliable moving instance segmenta- tion in sparse radar point clouds.IEEE Transactions on Robotics, 40:2357–2372, 2023

    Matthias Zeller, Vardeep S Sandhu, Benedikt Mersch, Jens Behley, Michael Heidingsfeld, and Cyrill Stachniss. Radar instance transformer: Reliable moving instance segmenta- tion in sparse radar point clouds.IEEE Transactions on Robotics, 40:2357–2372, 2023. 1

    2023

  55. [55]

    Bo Zhang and Jian Zhang. A traffic surveillance system for obtaining comprehensive information of the passing vehicles based on instance segmentation.IEEE Transactions on In- telligent Transportation Systems, 22(11):7040–7055, 2020. 1

    2020

  56. [56]

    Adaptive multi-source predictor for zero-shot video object segmentation.International Journal of Computer Vision, 132(8):3232–3250, 2024

    Xiaoqi Zhao, Shijie Chang, Youwei Pang, Jiaxing Yang, Lihe Zhang, and Huchuan Lu. Adaptive multi-source predictor for zero-shot video object segmentation.International Journal of Computer Vision, 132(8):3232–3250, 2024. 2

    2024

  57. [57]

    Matnet: Motion-attentive transition network for zero-shot video object segmentation.IEEE transactions on image processing, 29:8326–8338, 2020

    Tianfei Zhou, Jianwu Li, Shunzhou Wang, Ran Tao, and Jianbing Shen. Matnet: Motion-attentive transition network for zero-shot video object segmentation.IEEE transactions on image processing, 29:8326–8338, 2020. 2

    2020

  58. [58]

    Event-based motion segmentation with spatio- temporal graph cuts.IEEE transactions on neural networks and learning systems, 34(8):4868–4880, 2021

    Yi Zhou, Guillermo Gallego, Xiuyuan Lu, Siqi Liu, and Shaojie Shen. Event-based motion segmentation with spatio- temporal graph cuts.IEEE transactions on neural networks and learning systems, 34(8):4868–4880, 2021. 6

    2021