Audio Spoofing Verification using Deep Convolutional Neural Networks by Transfer Learning
This addresses security vulnerabilities in speaker verification systems against replay attacks, but it is incremental as it adapts an existing ResNet architecture.
The paper tackled the problem of detecting spoofing attacks in automatic speaker verification systems by proposing a deep-convolutional neural network classifier, achieving an equal error rate as low as 0.9056% on development data and up to 5.87% on evaluation datasets.
Automatic Speaker Verification systems are gaining popularity these days; spoofing attacks are of prime concern as they make these systems vulnerable. Some spoofing attacks like Replay attacks are easier to implement but are very hard to detect thus creating the need for suitable countermeasures. In this paper, we propose a speech classifier based on deep-convolutional neural network to detect spoofing attacks. Our proposed methodology uses acoustic time-frequency representation of power spectral densities on Mel frequency scale (Mel-spectrogram), via deep residual learning (an adaptation of ResNet-34 architecture). Using a single model system, we have achieved an equal error rate (EER) of 0.9056% on the development and 5.32% on the evaluation dataset of logical access scenario and an equal error rate (EER) of 5.87% on the development and 5.74% on the evaluation dataset of physical access scenario of ASVspoof 2019.