A Holistic Cascade System, benchmark, and Human Evaluation Protocol for Expressive Speech-to-Speech Translation

TL;DR

This paper introduces a comprehensive cascade system for expressive speech-to-speech translation, along with a new benchmark dataset and human evaluation protocol, demonstrating improved expressive preservation over single-aspect methods.

Contribution

The work presents a holistic cascade approach, curated benchmark datasets, and a human evaluation protocol for expressive S2ST, addressing gaps in evaluation and multi-aspect expressivity modeling.

Findings

Holistic cascade system outperforms single-aspect models.

Bi-lingual annotators effectively evaluate expressive quality.

Benchmark datasets in TV series and audiobook domains.

Abstract

Expressive speech-to-speech translation (S2ST) aims to transfer prosodic attributes of source speech to target speech while maintaining translation accuracy. Existing research in expressive S2ST is limited, typically focusing on a single expressivity aspect at a time. Likewise, this research area lacks standard evaluation protocols and well-curated benchmark datasets. In this work, we propose a holistic cascade system for expressive S2ST, combining multiple prosody transfer techniques previously considered only in isolation. We curate a benchmark expressivity test set in the TV series domain and explored a second dataset in the audiobook domain. Finally, we present a human evaluation protocol to assess multiple expressive dimensions across speech pairs. Experimental results indicate that bi-lingual annotators can assess the quality of expressive preservation in S2ST systems, and the holistic modeling approach outperforms single-aspect systems. Audio samples can be accessed through our demo webpage: https://facebookresearch.github.io/speech_translation/cascade_expressive_s2st.