Program Description

We collect surveys or brief technical contributions about the privacy-preserving surveillance systems (PPSS). Authors are expected to submit extended abstract of the paper containing more than 300 words for each submission. If their contributions were accepted, the authors will give a 15-minute oral presentation at the workshop. The accepted contributions will be combined into a single merged workshop paper which will be included and published to the proceedings of IEEE AVSS 2023. Abstracts submitted to the merged survey will be considered for publication of the workshop paper after editorial modifications.

Workshop Topics

We conduct an extensive investigation of privacy-protecting methodologies for the surveillance and monitoring applications. Moreover, we explore not only direct information protection but also theoretical protection methods against learning model attacks such as inversion attacks, membership inference attacks. We will invite speakers with professional knowledge and experience about the topics:

• Privacy protection methods for 2D/3D visual data included RGB images, 2D/3D key points or features, point clouds, and others

• Privacy threats or protection methods for large language models

• Privacy-preserving LiDAR, RADAR image/signal applications

• Privacy-preserving voice signal applications

• Theoretical foundations against model inversion attacks, membership inference attack such as differential privacy, homomorphic encryption and, federated learning.

• Hardware based privacy-preserving techniques in the point of optics or circuits.

Importants Dates

Contribution Paper Deadline: September 10 23:59, 2023

Acceptance Notification Date: October 1 23:59, 2023

Camera Ready Submission: October 11 23:59, 2023

Workshop Date: November 6 23:59, 2023

Please check the pacific time zone, UTC+09:00.