Similarity Learning with Top-heavy Ranking Loss for Person Re-identification

PROJECT TITLE :

Similarity Learning with Top-heavy Ranking Loss for Person Re-identification

ABSTRACT:

Person re-identification is that the task of finding a person of interest across a network of cameras. In this paper, we have a tendency to propose a new similarity learning technique for person re-identification. Standard metric learning ways generally learn a linear transformation by using sparse pairwise or triplet constraints. Since a ton of negative matching pairs or triplets are abandoned, the discriminative information isn't absolutely exploited. Similarity learning methods with AUC loss can utilize all valid triplet constraints. However, the AUC loss has its own limitation by treating all false ranks occured at different positions equally. To handle this limitation, we have a tendency to propose to extend the AUC loss to the top-heavy ranking loss by assigning large weights to top positions of the ranking list. Moreover, we have a tendency to introduce an specific nonlinear transformation perform for the first feature space and learn an inner product similarity underneath the structured output learning framework. Our approach achieves terribly promising results on the challenging VIPeR, CUHK Campus and PRID 450S datasets.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here