arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Computer Science · Computer Vision and Pattern Recognition

Appearance-based people recognition by local dissimilarity representations

Riccardo Satta, Giorgio Fumera, Fabio Roli

· UNICA IRIS Institutional Research Information System (University of Cagliari)

See on the map

Abstract

Among the possible applications of computer vision to video-surveillance, person re-identification over a network of camera sensors, using cues related to clothing appearance, is gaining much interest. Re-identification techniques can be used for various tasks, e.g., online tracking of a person, and off-line retrieval of all video sequences containing an individual of interest, whose image is given as a query. Recently, some authors proposed to exploit clothing appearance descriptors also to retrieve video sequences of individuals that match a textual description of clothing (e.g., "person wearing a black t-shirt and white trousers"), instead of an image. We name this task "appearance-based people search". This functionality can be useful, e.g., in forensics investigations, where a textual description can be provided by a witness. In this paper, we present and experimentally evaluate a general method to perform both person re-identification and people search, using any given descriptor of clothing appearance that exploits widely used multiple part/multiple component representations. It is based on turning the considered appearance descriptor into a dissimilarity-based one, through a framework we previously proposed for speeding up person re-identification methods. Our approach allows one to deploy systems able to perform both tasks with the same pipeline and processing stages for constructing descriptors.

More like this