Person Re-identification with Bias-controlled Adversarial Training

Sara Iodice, Krystian Mikolajczyk

Abstract

Inspired by the effectiveness of adversarial training in the area of Generative Adversarial Networks we present a new approach for learning feature representations in person re-identification. We investigate different types of bias that typically occur in re-ID scenarios, i.e., pose, body part and camera view, and propose a general approach to address them. We introduce an adversarial strategy for controlling bias, named Bias-controlled Adversarial framework (BCA), with two complementary branches to reduce or to enhance bias-related features. The results and comparison to the state of the art on different benchmarks show that our framework is an effective strategy for person re-identification. The performance improvements are in both full and partial views of persons.

Results

TaskDatasetMetricValueModel
Person Re-IdentificationMarket-1501Rank-193.1Bias-controlled Adversarial Training
Person Re-IdentificationMarket-1501mAP89.3Bias-controlled Adversarial Training
Person Re-IdentificationDukeMTMC-reIDRank-185.2Bias-controlled Adversarial Training
Person Re-IdentificationDukeMTMC-reIDmAP74.8Bias-controlled Adversarial Training

Related Papers