Full Body Anonymization
Full-body anonymization aims to remove personally identifiable information from images and videos while preserving contextual information like pose and movement. Current research heavily utilizes generative adversarial networks (GANs), often enhanced with techniques like style-based generation and surface-guided synthesis, to create realistic replacements for individuals' appearances. This field is crucial for balancing privacy concerns with the continued use of visual data in various applications, including industrial settings and social media, by enabling the ethical use of data for research and development while protecting individual identities.
Papers
October 11, 2024
August 6, 2024
June 13, 2024
May 29, 2024
November 17, 2022
November 2, 2022