GMM-ResNet2: Ensemble of Group ResNet Networks for Synthetic Speech Detection

TL;DR

GMM-ResNet2 introduces an ensemble of group ResNet networks with multi-scale GMM features and an ensemble-aware loss for improved synthetic speech detection, outperforming previous models on benchmark datasets.

Contribution

The paper presents GMM-ResNet2, a novel ensemble model with multi-scale GMM features, group classification, and an ensemble-aware loss, advancing synthetic speech detection accuracy.

Findings

Achieves state-of-the-art results on ASVspoof 2019 and 2021 datasets.

Reduces EER and t-DCF significantly compared to baseline models.

Demonstrates the effectiveness of multi-scale GMM features and ensemble loss in detection tasks.

Abstract

Deep learning models are widely used for speaker recognition and spoofing speech detection. We propose the GMM-ResNet2 for synthesis speech detection. Compared with the previous GMM-ResNet model, GMM-ResNet2 has four improvements. Firstly, the different order GMMs have different capabilities to form smooth approximations to the feature distribution, and multiple GMMs are used to extract multi-scale Log Gaussian Probability features. Secondly, the grouping technique is used to improve the classification accuracy by exposing the group cardinality while reducing both the number of parameters and the training time. The final score is obtained by ensemble of all group classifier outputs using the averaging method. Thirdly, the residual block is improved by including one activation function and one batch normalization layer. Finally, an ensemble-aware loss function is proposed to integrate the independent loss functions of all ensemble members. On the ASVspoof 2019 LA task, the GMM-ResNet2 achieves a minimum t-DCF of 0.0227 and an EER of 0.79\%. On the ASVspoof 2021 LA task, the GMM-ResNet2 achieves a minimum t-DCF of 0.2362 and an EER of 2.19\%, and represents a relative reductions of 31.4\% and 76.3\% compared with the LFCC-LCNN baseline.