King's College London

Research portal

Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting

Research output: Contribution to journalArticlepeer-review

Standard

Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting. / Liu, Yuting; Wang, Zheng; Shi, Miaojing et al.

In: NEUROCOMPUTING, Vol. 494, 14.07.2022, p. 418-431.

Research output: Contribution to journalArticlepeer-review

Harvard

Liu, Y, Wang, Z, Shi, M, Satoh, S, Zhao, Q & Yang, H 2022, 'Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting', NEUROCOMPUTING, vol. 494, pp. 418-431. https://doi.org/10.1016/j.neucom.2022.04.107

APA

Liu, Y., Wang, Z., Shi, M., Satoh, S., Zhao, Q., & Yang, H. (2022). Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting. NEUROCOMPUTING, 494, 418-431. https://doi.org/10.1016/j.neucom.2022.04.107

Vancouver

Liu Y, Wang Z, Shi M, Satoh S, Zhao Q, Yang H. Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting. NEUROCOMPUTING. 2022 Jul 14;494:418-431. https://doi.org/10.1016/j.neucom.2022.04.107

Author

Liu, Yuting ; Wang, Zheng ; Shi, Miaojing et al. / Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting. In: NEUROCOMPUTING. 2022 ; Vol. 494. pp. 418-431.

Bibtex Download

@article{7e6d7700984d45749b113f945f79840d,
title = "Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting",
abstract = "Despite impressive progress in crowd counting over the last years, it is still an open challenge to reliably count crowds across visual domains. This paper addresses this setting, presenting an unsupervised cross-domain crowd counting framework able to perform unsupervised adaptation across domains with available unlabeled target data. We achieve this by learning to discover bi-knowledge transfer between regression- and detection-based models from a labeled source domain. The dual source knowledge of the two models is heterogeneous and complementary as they capture different modalities of crowd distribution. Specifically, we start by formulating the mutual transformations between the outputs of regression- and detection-based models as two scene-agnostic transformers which enable knowledge transfer between the two models. Given the regression- and detection-based models and their mutual transformers learnt on the source, we then introduce a self-supervised co-training scheme to encourage the knowledge transfer between the two models on the target. We further enhance the model adaptation with our modified mixup augmentation strategy. A thorough benchmark analysis against the most recent cross-domain crowd counting methods and detailed ablation studies show the advantage of our method.",
author = "Yuting Liu and Zheng Wang and Miaojing Shi and Shin'ichi Satoh and Qijun Zhao and Hongyu Yang",
note = "Funding Information: The research was partly supported by National Natural Science Foundation of China (No. 62176170, 61828602, 62066042, 61971005), JST CREST Grant (JPMJCR1686), and Grant-in-Aid for JSPS Fellows (18F18378). Publisher Copyright: {\textcopyright} 2022",
year = "2022",
month = jul,
day = "14",
doi = "10.1016/j.neucom.2022.04.107",
language = "English",
volume = "494",
pages = "418--431",
journal = "NEUROCOMPUTING",
issn = "0925-2312",
publisher = "Elsevier",

}

RIS (suitable for import to EndNote) Download

TY - JOUR

T1 - Discovering Regression-Detection Bi-knowledge Transfer for Unsupervised Cross-Domain Crowd Counting

AU - Liu, Yuting

AU - Wang, Zheng

AU - Shi, Miaojing

AU - Satoh, Shin'ichi

AU - Zhao, Qijun

AU - Yang, Hongyu

N1 - Funding Information: The research was partly supported by National Natural Science Foundation of China (No. 62176170, 61828602, 62066042, 61971005), JST CREST Grant (JPMJCR1686), and Grant-in-Aid for JSPS Fellows (18F18378). Publisher Copyright: © 2022

PY - 2022/7/14

Y1 - 2022/7/14

N2 - Despite impressive progress in crowd counting over the last years, it is still an open challenge to reliably count crowds across visual domains. This paper addresses this setting, presenting an unsupervised cross-domain crowd counting framework able to perform unsupervised adaptation across domains with available unlabeled target data. We achieve this by learning to discover bi-knowledge transfer between regression- and detection-based models from a labeled source domain. The dual source knowledge of the two models is heterogeneous and complementary as they capture different modalities of crowd distribution. Specifically, we start by formulating the mutual transformations between the outputs of regression- and detection-based models as two scene-agnostic transformers which enable knowledge transfer between the two models. Given the regression- and detection-based models and their mutual transformers learnt on the source, we then introduce a self-supervised co-training scheme to encourage the knowledge transfer between the two models on the target. We further enhance the model adaptation with our modified mixup augmentation strategy. A thorough benchmark analysis against the most recent cross-domain crowd counting methods and detailed ablation studies show the advantage of our method.

AB - Despite impressive progress in crowd counting over the last years, it is still an open challenge to reliably count crowds across visual domains. This paper addresses this setting, presenting an unsupervised cross-domain crowd counting framework able to perform unsupervised adaptation across domains with available unlabeled target data. We achieve this by learning to discover bi-knowledge transfer between regression- and detection-based models from a labeled source domain. The dual source knowledge of the two models is heterogeneous and complementary as they capture different modalities of crowd distribution. Specifically, we start by formulating the mutual transformations between the outputs of regression- and detection-based models as two scene-agnostic transformers which enable knowledge transfer between the two models. Given the regression- and detection-based models and their mutual transformers learnt on the source, we then introduce a self-supervised co-training scheme to encourage the knowledge transfer between the two models on the target. We further enhance the model adaptation with our modified mixup augmentation strategy. A thorough benchmark analysis against the most recent cross-domain crowd counting methods and detailed ablation studies show the advantage of our method.

UR - http://www.scopus.com/inward/record.url?scp=85129607306&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2022.04.107

DO - 10.1016/j.neucom.2022.04.107

M3 - Article

VL - 494

SP - 418

EP - 431

JO - NEUROCOMPUTING

JF - NEUROCOMPUTING

SN - 0925-2312

ER -

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454