eprintid: 4260
rev_number: 11
eprint_status: archive
userid: 411
dir: disk0/00/00/42/60
datestamp: 2020-12-14 05:02:45
lastmod: 2020-12-14 05:03:05
status_changed: 2020-12-14 05:02:45
type: conference_item
metadata_visibility: show
creators_name: Vu, Xuan Son
creators_name: Le, Duc Trong
creators_name: Edlund, Christoffer
creators_name: Jiang, Lili
creators_name: Nguyen, Hoang D.
creators_id: sonvx@cs.umu.se
creators_id: trongld@vnu.edu.vn
creators_id: christoffer.edlund@sartorius.com
creators_id: lili.jiang@cs.umu.se
creators_id: Harry.Nguyen@glasgow.ac.uk
title: Privacy-Preserving Visual Content Tagging using Graph Transformer Networks
ispublished: pub
subjects: IT
subjects: isi_scopus_conf
divisions: fac_fit
abstract: With the rapid growth of Internet media, content tagging has become an important topic with many multimedia understanding applications, including efficient organisation and search. Nevertheless, existing visual tagging approaches are susceptible to inherent privacy risks in which private information may be exposed unintentionally. The use of anonymisation and privacy-protection methods is desirable, but with the expense of task performance.Therefore, this paper proposes an end-to-end framework (SGTN) using Graph Transformer and Convolutional Networks to significantly improve classification and privacy preservation of visual data. Especially, we employ several mechanisms such as differential privacy based graph construction and noise-induced graph transformation to protect the privacy of knowledge graphs. Our approach unveils new state-of-the-art on MS-COCO dataset in various semisupervised settings. In addition, we showcase a real experiment in the education domain to address the automation of sensitive document tagging. Experimental results show that our approach achieves an excellent balance of model accuracy and privacy preservation on both public and private datasets. Codes are available at https://github.com/ReML-AI/sgtn.
date: 2020-10
date_type: published
official_url: https://doi.org/10.1145/3394171.3414047
full_text_status: public
pres_type: paper
event_title: The 28th ACM International Conference on Multimedia
event_type: conference
refereed: TRUE
citation:   Vu, Xuan Son and Le, Duc Trong and Edlund, Christoffer and Jiang, Lili and Nguyen, Hoang D.  (2020) Privacy-Preserving Visual Content Tagging using Graph Transformer Networks.  In: The 28th ACM International Conference on Multimedia.     
document_url: https://eprints.uet.vnu.edu.vn/eprints/id/eprint/4260/1/2020_ACMM.pdf