HPILN: a feature learning framework for cross‐modality person re‐identification. Issue 14 (23rd October 2019)

• Record Type: Journal Article
• Authors: Zhao, Yun‐Bo; Lin, Jian‐Wu; Xuan, Qi; Xi, Xugang
• Is Part Of: IET image processing. Volume 13:Issue 14(2019)
• Page Start: 2897
• Page End: 2904
• Publication Date: 2019-10-23
• Subjects: learning (artificial intelligence) -- feature extraction -- image classification -- video surveillance -- cameras -- pose estimation -- image colour analysis; modified cross‐modality re‐identification models -- HPILN -- feature learning framework -- cross‐modality person -- video surveillance systems -- RGB -- cameras -- cross‐modality variations -- heterogeneous images -- intra‐modality variations -- heterogeneous human poses -- camera position -- identity loss network -- single‐modality re‐identification models -- cross‐modality scenario -- SYSU‐MM01 dataset -- cumulative match characteristic curve -- mean average precision; Image processing -- Periodicals; 621.36705
• DOI: 10.1049/iet-ipr.2019.0699
• Languages: English
• ISSNs: 1751-9659

Abstract

Abstract : Most video surveillance systems use both RGB and infrared cameras, making it a vital technique to re‐identify a person cross the RGB and infrared modalities. This task can be challenging due to both the cross‐modality variations caused by heterogeneous images in RGB and infrared, and the intra‐modality variations caused by the heterogeneous human poses, camera position, light brightness etc. To meet these challenges, a novel feature learning framework, hard pentaplet and identity loss network (HPILN), is proposed. In the framework existing single‐modality re‐identification models are modified to fit for the cross‐modality scenario, following which specifically designed hard pentaplet loss and identity loss are used to increase the accuracy of the modified cross‐modality re‐identification models. Based on the benchmark of the SYSU‐MM01 dataset, extensive experiments have been conducted, showing that the authors' method outperforms all existing ones in terms of cumulative match characteristic curve and mean average precision.