Curricular SincNet: Towards Robust Deep Speaker Recognition by Emphasizing Hard Samples in Latent Space

TL;DR

This paper introduces Curricular SincNet, a deep speaker recognition model that emphasizes hard samples in latent space using a curriculum loss, leading to improved robustness and lower error rates across multiple datasets.

Contribution

The paper proposes a novel curriculum loss function for SincNet, enhancing speaker recognition by focusing on hard samples and imposing inter-class margins.

Findings

Achieves best overall results in inter-dataset testing with 4% lower error rate.

Performs competitively on multiple datasets in intra- and inter-dataset evaluations.

Outperforms previous SincNet models and published work in robustness and accuracy.

Abstract

Deep learning models have become an increasingly preferred option for biometric recognition systems, such as speaker recognition. SincNet, a deep neural network architecture, gained popularity in speaker recognition tasks due to its parameterized sinc functions that allow it to work directly on the speech signal. The original SincNet architecture uses the softmax loss, which may not be the most suitable choice for recognition-based tasks. Such loss functions do not impose inter-class margins nor differentiate between easy and hard training samples. Curriculum learning, particularly those leveraging angular margin-based losses, has proven very successful in other biometric applications such as face recognition. The advantage of such a curriculum learning-based techniques is that it will impose inter-class margins as well as taking to account easy and hard samples. In this paper, we propose Curricular SincNet (CL-SincNet), an improved SincNet model where we use a curricular loss function to train the SincNet architecture. The proposed model is evaluated on multiple datasets using intra-dataset and inter-dataset evaluation protocols. In both settings, the model performs competitively with other previously published work. In the case of inter-dataset testing, it achieves the best overall results with a reduction of 4\% error rate compare to SincNet and other published work.