This paper describes NAIST's submission to the simultaneous track of the IWSLT 2024 Evaluation Campaign: English-to-{German, Japanese, Chinese} speech-to-text translation and English-to-Japanese speech-to-speech translation. We develop a multilingual end-to-end speech-to-text translation model combining two pre-trained language models, HuBERT and mBART. We trained this model with two decoding policies, Local Agreement (LA) and AlignAtt. The submitted models employ the LA policy because it outperformed the AlignAtt policy in previous models. Our speech-to-speech translation method is a cascade of the above speech-to-text model and an incremental text-to-speech (TTS) module that incorporates a phoneme estimation model, a parallel acoustic model, and a parallel WaveGAN vocoder. We improved our incremental TTS by applying the Transformer architecture with the AlignAtt policy for the estimation model. The results show that our upgraded TTS module contributed to improving the system performance.
@article{arxiv.2407.00826,
title = {NAIST Simultaneous Speech Translation System for IWSLT 2024},
author = {Yuka Ko and Ryo Fukuda and Yuta Nishikawa and Yasumasa Kano and Tomoya Yanagita and Kosuke Doi and Mana Makinae and Haotian Tan and Makoto Sakai and Sakriani Sakti and Katsuhito Sudoh and Satoshi Nakamura},
journal= {arXiv preprint arXiv:2407.00826},
year = {2024}
}