CoChat
Sign Up
Cited by 50Video Surveillance and Tracking MethodsAdvanced Neural Network ApplicationsAdvanced Image and Video Retrieval Techniques

The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.

Content-Adaptive Auto-Occlusion Network for Occluded Person Re-Identification

Zhao, Cairong, Qu, Zefan, Jiang, Xinyang, Tu, Yuanpeng, Bai, Xiang·2023·IEEE Transactions on Image Processing
DOI 10.1109/tip.2023.3290525
Open this paper in CoChat — free account

Abstract

The occluded person re-identification (ReID) aims to match person images captured in severely occluded environments. Current occluded ReID works mostly rely on auxiliary models or employ a part-to-part matching strategy. However, these methods may be sub-optimal since the auxiliary models are constrained by occlusion scenes and the matching strategy will deteriorate when both query and gallery set contain occlusion. Some methods attempt to solve this problem by applying image occlusion augmentation (OA) and have shown great superiority in their effectiveness and lightness. But there are two defects that existed in the previous OA-based method: 1) The occlusion policy is fixed throughout the entire training and cannot be dynamically adjusted based on the current training status of the ReID network. 2) The position and area of the applied OA are completely random, without reference to the image content to choose the most suitable policy. To address these challenges, we propose a novel Content-Adaptive Auto-Occlusion Network (CAAO), that is able to dynamically select the proper occlusion region of an image based on its content and the current training status. Specifically, CAAO consists of two parts: the ReID network and the Auto-Occlusion Controller (AOC) module. AOC automatically generates the optimal OA policy based on the feature map extracted from the ReID network and applies occlusion on the images for ReID network training. An on-policy reinforcement learning based altern…

Overview

Integrity signals

Registry

Crossref · 10.1109/tip.2023.3290525

Cites

2 of 2 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 53

  1. 10.1109/ICCV48922.2021.01162 · DOI
  2. 10.1109/TIP.2021.3070182 · DOI
  3. 10.1609/aaai.v36i3.20155 · DOI
  4. 10.1109/CVPR52688.2022.00471 · DOI
  5. 10.1109/ICCV.2019.00046 · DOI
  6. 10.1145/3240508.3240552 · DOI
  7. 10.1109/CVPR46437.2021.00292 · DOI
  8. 10.1109/ICCV.2019.00374 · DOI
  9. 10.1109/ICCV48922.2021.01166 · DOI
  10. 10.1109/CVPR.2018.00117 · DOI
  11. 10.1109/CVPR.2018.00535 · DOI
  12. 10.1109/CVPR.2019.00048 · DOI
  13. 10.1109/CVPR.2019.00020 · DOI
  14. 10.1145/3503161.3547764 · DOI
  15. 10.1109/ICCV.2015.531 · DOI
  16. 10.1609/aaai.v35i2.16260 · DOI
  17. he, 2018, Recognizing partial biometric patterns, arXiv 1810 07399
  18. 10.1007/978-3-030-01225-0_30 · DOI
  19. 10.1145/3474085.3475610 · DOI
  20. zhuo, 2019, A novel teacher-student learning framework for occluded person re-identification, arXiv 1907 03253
  21. 10.1109/CVPR.2018.00243 · DOI
  22. zhou, 2019, Torchreid: A library for deep learning person re-identification in pytorch, arXiv 1910 10093
  23. 10.1109/ICCV.2017.349 · DOI
  24. 10.1109/ICCV48922.2021.01474 · DOI · Paper page
  25. 10.1109/TMM.2022.3156282 · DOI
  26. 10.1109/CVPR.2018.00739 · DOI
  27. 10.1109/ICCV48922.2021.01167 · DOI
  28. 10.1109/ICCV.2019.00854 · DOI
  29. 10.1109/CVPR42600.2020.01176 · DOI
  30. 10.1109/CVPR42600.2020.00648 · DOI
  31. hou, 2022, Feature completion for occluded person re-identification, IEEE Trans Pattern Anal Mach Intell, 44, 4894
  32. ge, 2018, FD-GAN: Pose-guided feature distilling GAN for robust person re-identification, Proc NeurIPS, 1222
  33. 10.1109/ICCV.2017.405 · DOI
  34. devries, 2017, Improved regularization of convolutional neural networks with cutout, arXiv 1708 04552
  35. 10.1007/BF00992696 · DOI
  36. mnih, 2016, Asynchronous methods for deep reinforcement learning, Proc Int Conf Mach Learn, 1928
  37. 10.1109/CVPR52688.2022.01063 · DOI
  38. 10.1007/978-3-030-58580-8_21 · DOI
  39. puterman, 2014, Markov Decision Processes Discrete Stochastic Dynamic Programming
  40. 10.1109/ICCV.2019.00063 · DOI
  41. 10.1109/ICME.2018.8486568 · DOI
  42. 10.1007/978-3-319-48881-3_2 · DOI
  43. 10.1109/ICCV.2015.133 · DOI
  44. 10.1109/ICCV48922.2021.00081 · DOI
  45. 10.1109/CVPRW50498.2020.00359 · DOI · Paper page
  46. 10.1109/CVPR42600.2020.00641 · DOI
  47. 10.1007/978-3-030-58571-6_19 · DOI
  48. 10.1007/978-3-030-58595-2_1 · DOI
  49. li, 2020, DADA: Differentiable automatic data augmentation, arXiv 2003 03780
  50. lim, 2019, Fast autoaugment, Proc Adv Neural Inf Process Syst, 32, 1
  51. 10.1109/CVPR46437.2021.00111 · DOI
  52. zhang, 2019, Adversarial AutoAugment, arXiv 1912 11188
  53. 10.1609/aaai.v35i11.17127 · DOI

Cited by

Sorted by citations

Metadata

Authors
Zhao, C., Qu, Z., Jiang, X., Tu, Y., & Bai, X.
Affiliations
  • Tongji University (CN)
  • Microsoft Research Asia (Shanghai), Shanghai, China
  • Huazhong University of Science and Technology (CN)
ORCID
Venue
IEEE Transactions on Image Processing
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Published
2023
Volume
32
Pages
4223-4236
ISSN
  • 1057-7149 (ISSN-L)
  • 1941-0042
DOI
10.1109/tip.2023.3290525
OpenAlex
W4383200119
Cited by
50
References
2
Citations by year
2026: 18 · 2025: 19 · 2024: 13
FWCI
4.52
Type
journal article
Language
en
Open access
closed
Licence
Locations
  • IEEE Transactions on Image Processing (publishedVersion)
  • PubMed (publishedVersion)
Funders
  • National Natural Science Fund of China (62076184, 61976158, 61976160, 62076182, 62276190)
  • Fundamental Research Funds for the Central Universities
  • State Key Laboratory of Integrated Services Networks
  • Shanghai Innovation Action Project of Science and Technology (20511100700)
  • Shanghai Natural Science Foundation (22ZR1466700)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Video Surveillance and Tracking Methods
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
Keywords
occluded person re-identification, on-policy learning, alternating training

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review