Buenos Aires, Argentina
Chalearn LAP Workshop @FG2020 "Privacy-aware Computer Vision"
Aims and scope: Preserving people’s privacy is a crucial issue faced by many computer vision applications. While exploiting video data from RGB cameras has been proven successful in many problems, it may come at the higher cost of compromising observed individual’s sensitive data. This negatively affects the popularity of visual information systems despite their enormous potential to help people in their everyday’s life. Privacy-aware visual information systems need to minimize the amount of potentially sensitive information that is being collected and/or handled through their pipelines while still achieving a reliable performance. We aim to compile latest efforts and research advances from the scientific community in privacy preservation in computer vision and pattern recognition algorithms at data collection, learning, and decision stages.
Associated contest: We are organizing an associated competition on identity-preserving human detection on the more privacy-safe depth and thermal modalities. The participants will be invited to submit their papers to this workshop.
Workshop topics and guidelines: The scope of the workshop comprises all aspects of privacy preservation in computer vision methods for human-related analyses. Most notably, the following are the main topics of interest:
Unconventional but more privacy-safe visual modalities and/or fusion of those.
Novel identity-preserving methods in all domains of human analysis (e.g. human detection, tracking, pose recovery, re-identification, action/gesture/facial expression/scene recognition, and so on).
Novel methods for early detection of privacy-sensitive situations.
Novel methods for detection, removal, etc. of privacy-sensitive objects/scene parts.
Novel privacy-aware dataset papers.
Studies of the new advances on all aspects of privacy in computer vision.
New privacy-aware vision applications/systems.