How to Train Triplet Networks with 100K Identities?

TL;DR

This paper introduces a subspace learning approach combined with batch hard negative mining to effectively train triplet networks on large-scale face recognition datasets with 100K identities, improving performance and robustness.

Contribution

The paper proposes a novel subspace learning method for triplet networks that addresses the challenge of large-scale identity sets in face recognition.

Findings

Significant performance improvement on MS-Celeb-1M dataset.

Effective handling of noisy data and large-scale retrieval.

Achieved state-of-the-art results without external data.

Abstract

Training triplet networks with large-scale data is challenging in face recognition. Due to the number of possible triplets explodes with the number of samples, previous studies adopt the online hard negative mining(OHNM) to handle it. However, as the number of identities becomes extremely large, the training will suffer from bad local minima because effective hard triplets are difficult to be found. To solve the problem, in this paper, we propose training triplet networks with subspace learning, which splits the space of all identities into subspaces consisting of only similar identities. Combined with the batch OHNM, hard triplets can be found much easier. Experiments on the large-scale MS-Celeb-1M challenge with 100K identities demonstrate that the proposed method can largely improve the performance. In addition, to deal with heavy noise and large-scale retrieval, we also make some efforts on robust noise removing and efficient image retrieval, which are used jointly with the subspace learning to obtain the state-of-the-art performance on the MS-Celeb-1M competition (without external data in Challenge1).