Raw Differentiable Architecture Search for Speech Deepfake and Spoofing Detection

TL;DR

This paper introduces a method for automatically learning the architecture of speech deepfake and spoofing detection systems directly from raw audio signals, achieving state-of-the-art results.

Contribution

It presents a novel raw differentiable architecture search approach that jointly optimizes network architecture and parameters for speech anti-spoofing.

Findings

Achieved a tandem detection cost function score of 0.0517 on ASVspoof 2019 database

Outperforms many existing single-system approaches in anti-spoofing

Demonstrates the effectiveness of learned architectures over hand-crafted ones

Abstract

End-to-end approaches to anti-spoofing, especially those which operate directly upon the raw signal, are starting to be competitive with their more traditional counterparts. Until recently, all such approaches consider only the learning of network parameters; the network architecture is still hand crafted. This too, however, can also be learned. Described in this paper is our attempt to learn automatically the network architecture of a speech deepfake and spoofing detection solution, while jointly optimising other network components and parameters, such as the first convolutional layer which operates on raw signal inputs. The resulting raw differentiable architecture search system delivers a tandem detection cost function score of 0.0517 for the ASVspoof 2019 logical access database, a result which is among the best single-system results reported to date.