A speech spoofing countermeasure (CM) that discriminates between unseen spoofed and bona fide data requires diverse training data. While many datasets use spoofed data generated by speech synthesis systems, it was recently found that data vocoded by neural vocoders were also effective as the spoofed training data. Since many neural vocoders are fast in building and generation, this study used multiple neural vocoders and created more than 9,000 hours of vocoded data on the basis of the VoxCeleb2 corpus. This study investigates how this large-scale vocoded data can improve spoofing countermeasures that use data-hungry self-supervised learning (SSL) models. Experiments demonstrated that the overall CM performance on multiple test sets improved when using features extracted by an SSL model continually trained on the vocoded data. Further improvement was observed when using a new SSL distilled from the two SSLs before and after the continual training. The CM with the distilled SSL outperformed the previous best model on challenging unseen test sets, including the ASVspoof 2019 logical access, WaveFake, and In-the-Wild.

翻译：语音伪造检测技术（CM）需要多样化的训练数据以区分未见过的伪造语音与真实语音。尽管现有数据集多采用语音合成系统生成的伪造数据，但近期研究发现，神经声码器处理的声码化数据同样可作为有效的伪造训练数据。由于多数神经声码器构建与生成速度快，本研究基于VoxCeleb2语料库，利用多个神经声码器生成了超过9000小时的声码化数据。本文探究了这种大规模声码化数据如何改进依赖大规模数据的自监督学习（SSL）模型的伪造检测性能。实验表明，当使用持续在声码化数据上训练的SSL模型提取特征时，多个测试集上的整体CM性能得到提升。进一步，通过从持续训练前后的两个SSL模型中蒸馏出的新SSL模型，性能获得更大改善。采用蒸馏SSL的CM模型在具有挑战性的未见测试集（包括ASVspoof 2019逻辑访问、WaveFake和In-the-Wild）上超越了此前最优模型。