Improving learning objectives for speaker verification from the perspective of score comparison

Deep speaker embedding systems are usually trained with classification-based or end-to-end learning objectives. Popular end-to-end approaches utilize deep metric learning, which can be viewed as a few-shot classification objective. In this paper, we investigate the limit of conventional learning objectives in speaker verification, and propose a new learning objective designed from the perspective of similarity scores. The proposed method trains a network by score comparison unbound from the classification situation, which is more suitable for verification tasks. Experiments conducted with various network architectures demonstrate the improvements on the VoxCeleb dataset using the proposed loss.