Graph Attention Networks for Anti-Spoofing

Hemlata Tak,Jee-weon Jung,Jose Patino,Massimiliano Todisco,Nicholas Evans
DOI: https://doi.org/10.48550/arXiv.2104.03654
2021-04-08
Abstract:The cues needed to detect spoofing attacks against automatic speaker verification are often located in specific spectral sub-bands or temporal segments. Previous works show the potential to learn these using either spectral or temporal self-attention mechanisms but not the relationships between neighbouring sub-bands or segments. This paper reports our use of graph attention networks (GATs) to model these relationships and to improve spoofing detection performance. GATs leverage a self-attention mechanism over graph structured data to model the data manifold and the relationships between nodes. Our graph is constructed from representations produced by a ResNet. Nodes in the graph represent information either in specific sub-bands or temporal segments. Experiments performed on the ASVspoof 2019 logical access database show that our GAT-based model with temporal attention outperforms all of our baseline single systems. Furthermore, GAT-based systems are complementary to a set of existing systems. The fusion of GAT-based models with more conventional countermeasures delivers a 47% relative improvement in performance compared to the best performing single GAT system.
Audio and Speech Processing,Cryptography and Security,Sound
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the detection of spoofing attacks in Automatic Speaker Verification (ASV) systems. Specifically, the paper focuses on how to effectively identify processing artefacts generated by voice signal manipulation or synthesis. These artefacts are usually present in specific spectral sub - bands or time segments. In order to improve the detection performance of these spoofing attacks, the paper proposes to use Graph Attention Networks (GATs) to model the relationships between different sub - bands or time segments, thereby capturing more complex non - linear features and improving the accuracy of spoof detection. The main contribution of the paper lies in using GATs to model the spectral and temporal relationships. It operates on graph - structured data through the self - attention mechanism to better capture the relationships between nodes (i.e., information of specific sub - bands or time segments), thereby improving the performance of spoof detection. Experimental results show that the GAT - based method outperforms multiple baseline systems on the ASVspoof 2019 logical access database, and the performance can be further improved by fusing different systems.