Staff View: Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset

Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset

Source-Free Domain Adaptation (SFDA) is an important research topic in domains with data privacy concerns. Existing SFDA studies have successfully achieved domain adaptation without revealing source domain data, significantly reducing the possibility of privacy leaks. However, complete SFDA (cSFDA),...

Full description

Saved in:

Bibliographic Details
Main Authors:	Lee, Kyungchae, Tan, Ying Hui, Chuah, Joon Huang, Youn, Chan-Hyun
Format:	Article
Published:	Institute of Electrical and Electronics Engineers 2024
Subjects:	TK Electrical engineering. Electronics Nuclear engineering
Online Access:	http://eprints.um.edu.my/47092/ https://doi.org/10.1109/ACCESS.2024.3466226
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my.um.eprints.47092
record_format	eprints
spelling	my.um.eprints.470922024-11-22T04:45:14Z http://eprints.um.edu.my/47092/ Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset Lee, Kyungchae Tan, Ying Hui Chuah, Joon Huang Youn, Chan-Hyun TK Electrical engineering. Electronics Nuclear engineering Source-Free Domain Adaptation (SFDA) is an important research topic in domains with data privacy concerns. Existing SFDA studies have successfully achieved domain adaptation without revealing source domain data, significantly reducing the possibility of privacy leaks. However, complete SFDA (cSFDA), which does not disclose even the model weights of the source domain, has not yet been adequately addressed. Considering the rapidly advancing fields of techniques such as model inversion attacks, we believe that discussions on this cSFDA scenario should be conducted promptly. To perform domain adaptation without revealing both the weights and the data of the source domain, we redefine the domain adaptation process by decomposing it into two stages: information vector extraction and embedding information transfer. In this paper, we propose a novel Spectrogram Secure Domain Adaptation via Encrypted Vector Transfer (SeDA-EVT) pipeline for medical auscultation data, which is achieved by sequentially merging two processes above. We first demonstrated the effectiveness of our information extraction strategy through classification task performance evaluation, showing that our first phase is capable of producing information-rich embeddings. Next, by applying the embedding information transfer to a newly collected clinical lung sound data set from an ER environment, we verified that our proposed pipeline can transfer rich information to the target domain without revealing any source domain components. Institute of Electrical and Electronics Engineers 2024 Article PeerReviewed Lee, Kyungchae and Tan, Ying Hui and Chuah, Joon Huang and Youn, Chan-Hyun (2024) Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset. IEEE Access, 12. pp. 148201-148215. ISSN 2169-3536, DOI https://doi.org/10.1109/ACCESS.2024.3466226 <https://doi.org/10.1109/ACCESS.2024.3466226>. https://doi.org/10.1109/ACCESS.2024.3466226 10.1109/ACCESS.2024.3466226
institution	Universiti Malaya
building	UM Library
collection	Institutional Repository
continent	Asia
country	Malaysia
content_provider	Universiti Malaya
content_source	UM Research Repository
url_provider	http://eprints.um.edu.my/
topic	TK Electrical engineering. Electronics Nuclear engineering
spellingShingle	TK Electrical engineering. Electronics Nuclear engineering Lee, Kyungchae Tan, Ying Hui Chuah, Joon Huang Youn, Chan-Hyun Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
description	Source-Free Domain Adaptation (SFDA) is an important research topic in domains with data privacy concerns. Existing SFDA studies have successfully achieved domain adaptation without revealing source domain data, significantly reducing the possibility of privacy leaks. However, complete SFDA (cSFDA), which does not disclose even the model weights of the source domain, has not yet been adequately addressed. Considering the rapidly advancing fields of techniques such as model inversion attacks, we believe that discussions on this cSFDA scenario should be conducted promptly. To perform domain adaptation without revealing both the weights and the data of the source domain, we redefine the domain adaptation process by decomposing it into two stages: information vector extraction and embedding information transfer. In this paper, we propose a novel Spectrogram Secure Domain Adaptation via Encrypted Vector Transfer (SeDA-EVT) pipeline for medical auscultation data, which is achieved by sequentially merging two processes above. We first demonstrated the effectiveness of our information extraction strategy through classification task performance evaluation, showing that our first phase is capable of producing information-rich embeddings. Next, by applying the embedding information transfer to a newly collected clinical lung sound data set from an ER environment, we verified that our proposed pipeline can transfer rich information to the target domain without revealing any source domain components.
format	Article
author	Lee, Kyungchae Tan, Ying Hui Chuah, Joon Huang Youn, Chan-Hyun
author_facet	Lee, Kyungchae Tan, Ying Hui Chuah, Joon Huang Youn, Chan-Hyun
author_sort	Lee, Kyungchae
title	Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
title_short	Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
title_full	Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
title_fullStr	Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
title_full_unstemmed	Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset
title_sort	privacy aware feature level domain adaptation for training deep vision models with private medical stethoscope dataset
publisher	Institute of Electrical and Electronics Engineers
publishDate	2024
url	http://eprints.um.edu.my/47092/ https://doi.org/10.1109/ACCESS.2024.3466226
_version_	1817841974218063872
score	13.223943

Privacy Aware Feature Level Domain Adaptation for Training Deep Vision Models With Private Medical Stethoscope Dataset

Similar Items