eprintid: 10115347 rev_number: 14 eprint_status: archive userid: 608 dir: disk0/10/11/53/47 datestamp: 2020-11-18 13:22:30 lastmod: 2021-12-16 01:43:29 status_changed: 2020-11-18 13:22:30 type: proceedings_section metadata_visibility: show creators_name: Ponce-Lopez, V creators_name: Burghardt, T creators_name: Sun, Y creators_name: Hannuna, S creators_name: Damen, D creators_name: Mirmehdi, M title: Deep Compact Person Re-Identification with Distractor Synthesis via Guided DC-GANs ispublished: pub divisions: UCL divisions: B04 divisions: C04 divisions: F34 keywords: Person Re-ID, GANs, Distractor synthesis, Deep face analysis note: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. abstract: We present a dual-stream CNN that learns both appearance and facial features in tandem from still images and, after feature fusion, infers person identities. We then describe an alternative architecture of a single, lightweight ID-CondenseNet where a face detector-guided DC-GAN is used to generate distractor person images for enhanced training. For evaluation, we test both architectures on FLIMA, a new extension of an existing person re-identification dataset with added frame-by-frame annotations of face presence. Although the dual-stream CNN can outperform the CondenseNet approach on FLIMA, we show that the latter surpasses all state-of-the-art architectures in top-1 ranking performance when applied to the largest existing person re-identification dataset, MSMT17. We conclude that whilst re-identification performance is highly sensitive to the structure of datasets, distractor augmentation and network compression have a role to play for enhancing performance characteristics for larger scale applications. date: 2019-09-02 date_type: published publisher: Springer official_url: https://doi.org/10.1007/978-3-030-30642-7_44 oa_status: green full_text_type: other language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 1818993 doi: 10.1007/978-3-030-30642-7_44 isbn_13: 978-3-030-30641-0 medium: 10.1007/978-3-030-30642-7_44 lyricists_name: Ponce Lopez, Victor lyricists_id: VPONC48 actors_name: Dewerpe, Marie actors_id: MDDEW97 actors_role: owner full_text_status: public series: Lecture Notes in Computer Science publication: IMAGE ANALYSIS AND PROCESSING - ICIAP 2019, PT I volume: 11751 place_of_pub: Cham, Switzerland pagerange: 488-498 pages: 11 event_title: 20th International Conference on Image Analysis and Processing (ICIAP) event_location: Univ Trento, Fac Law, Trento, ITALY event_dates: 09 September 2019 - 13 September 2019 institution: 20th International Conference on Image Analysis and Processing (ICIAP) issn: 1611-3349 book_title: Image Analysis and Processing – ICIAP 2019. ICIAP 2019 editors_name: Ricci, E editors_name: Bulo, SR editors_name: Snoek, C editors_name: Lanz, O editors_name: Messelodi, S editors_name: Sebe, N citation: Ponce-Lopez, V; Burghardt, T; Sun, Y; Hannuna, S; Damen, D; Mirmehdi, M; (2019) Deep Compact Person Re-Identification with Distractor Synthesis via Guided DC-GANs. In: Ricci, E and Bulo, SR and Snoek, C and Lanz, O and Messelodi, S and Sebe, N, (eds.) Image Analysis and Processing – ICIAP 2019. ICIAP 2019. (pp. pp. 488-498). Springer: Cham, Switzerland. Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10115347/1/Ponce%20Lopez_ICIAP19_Guided-DC-GANs_CameraReady.pdf