@inproceedings{discovery10115347, series = {Lecture Notes in Computer Science}, address = {Cham, Switzerland}, month = {September}, editor = {E Ricci and SR Bulo and C Snoek and O Lanz and S Messelodi and N Sebe}, year = {2019}, journal = {IMAGE ANALYSIS AND PROCESSING - ICIAP 2019, PT I}, publisher = {Springer}, volume = {11751}, note = {This version is the author accepted manuscript. For information on re-use, please refer to the publisher's terms and conditions.}, pages = {488--498}, booktitle = {Image Analysis and Processing - ICIAP 2019. ICIAP 2019}, title = {Deep Compact Person Re-Identification with Distractor Synthesis via Guided DC-GANs}, issn = {1611-3349}, keywords = {Person Re-ID, GANs, Distractor synthesis, Deep face analysis}, 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.}, url = {https://doi.org/10.1007/978-3-030-30642-7\%5f44}, author = {Ponce-Lopez, V and Burghardt, T and Sun, Y and Hannuna, S and Damen, D and Mirmehdi, M} }