Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification

Person re-identification (re-ID) models trained on one domain often fail to generalize well to another. In our attempt, we present a "learning via translation" framework. In the baseline, we translate the labeled images from source to target domain in an unsupervised manner. We then train re-ID models with the translated images by supervised methods. Yet, being an essential part of this framework, unsupervised image-image translation suffers from the information loss of source-domain labels during translation. Our motivation is two-fold. First, for each image, the discriminative cues contained in its ID label should be maintained after translation. Second, given the fact that two domains have entirely different persons, a translated image should be dissimilar to any of the target IDs. To this end, we propose to preserve two types of unsupervised similarities, 1) self-similarity of an image before and after translation, and 2) domain-dissimilarity of a translated source image and a target image. Both constraints are implemented in the similarity preserving generative adversarial network (SPGAN) which consists of an Siamese network and a CycleGAN. Through domain adaptation experiment, we show that images generated by SPGAN are more suitable for domain adaptation and yield consistent and competitive re-ID accuracy on two large-scale datasets.

PDF Abstract CVPR 2018 PDF CVPR 2018 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Person Re-Identification DukeMTMC-reID SPGAN+LMP* Rank-1 46.4 # 82
mAP 26.2 # 87
Unsupervised Person Re-Identification DukeMTMC-reID SPGAN+LMP Rank-1 46.4 # 11
Rank-10 68.0 # 10
Rank-5 62.3 # 10
MAP 26.2 # 11
Unsupervised Domain Adaptation Duke to Market SPGAN mAP 22.8 # 25
rank-1 51.5 # 24
rank-5 70.1 # 17
rank-10 76.8 # 17
Unsupervised Person Re-Identification Market-1501 SPGAN+LMP Rank-1 57.7 # 22
MAP 26.7 # 21
Rank-10 82.4 # 18
Rank-5 75.8 # 18
Unsupervised Domain Adaptation Market to Duke SPGAN mAP 22.3 # 24
rank-1 41.1 # 23
rank-5 56.6 # 16
rank-10 63.0 # 16
Unsupervised Person Re-Identification MSMT17->DukeMTMC-reID SPGAN Rank-1 46.4 # 3
Rank-10 68.0 # 3
Rank-5 62.3 # 3
mAP 26.2 # 3

Methods