The HCCL-DKU system for fake audio generation task of the 2022 ICASSP ADD Challenge
This work addresses the problem of generating convincing fake audio for spoofing anti-spoofing systems, representing an incremental advance in voice conversion technology.
The paper tackled fake audio generation for voice conversion by proposing a novel ppg-based fully end-to-end model, which outperformed existing models on conversion quality and spoofing performance, achieving second place in the 2022 ICASSP ADD Challenge with a deception success rate of 0.916.
The voice conversion task is to modify the speaker identity of continuous speech while preserving the linguistic content. Generally, the naturalness and similarity are two main metrics for evaluating the conversion quality, which has been improved significantly in recent years. This paper presents the HCCL-DKU entry for the fake audio generation task of the 2022 ICASSP ADD challenge. We propose a novel ppg-based voice conversion model that adopts a fully end-to-end structure. Experimental results show that the proposed method outperforms other conversion models, including Tacotron-based and Fastspeech-based models, on conversion quality and spoofing performance against anti-spoofing systems. In addition, we investigate several post-processing methods for better spoofing power. Finally, we achieve second place with a deception success rate of 0.916 in the ADD challenge.