Multiple-Shot Person Re-Identification by HPE Signature

Abstract--In this paper, we propose a novel appearancebased method for person re-identification, that condenses a set of frames of the same individual into a highly informative signature, called Histogram Plus Epitome, HPE. It incorporates complementary global and local statistical descriptions of the human appearance, focusing on the overall chromatic content, via histograms representation, and on the presence of recurrent local patches, via epitome estimation. The matching of HPEs provides optimal performances against low resolution, occlusions, pose and illumination variations, defining novel state-of-the-art results on all the datasets considered. Keywords-Pedestrian Re-identification, Pedestrian Description, Video-surveillance.

• Type: Conference
• Year: 2010
• Where: ICPR
• Authors: Loris Bazzani, Marco Cristani, Alessandro Perina, Michela Farenzena, Vittorio Murino
• Keywords: Computer Vision, Histogram Plus Epitome, ICPR 2010, Overall Chromatic Content, Recurrent Local Patches