Accent conversion using discrete units with parallel data synthesized from controllable accented TTS
This work addresses accent conversion for non-native speakers, offering a more flexible and efficient method compared to previous approaches.
The paper tackles accent conversion by converting non-native accents to native ones while preserving content and speaker identity, using discrete units and synthesized data to achieve improved fluency and native-like sound.
The goal of accent conversion (AC) is to convert speech accents while preserving content and speaker identity. Previous methods either required reference utterances during inference, did not preserve speaker identity well, or used one-to-one systems that could only be trained for each non-native accent. This paper presents a promising AC model that can convert many accents into native to overcome these issues. Our approach utilizes discrete units, derived from clustering self-supervised representations of native speech, as an intermediary target for accent conversion. Leveraging multi-speaker text-to-speech synthesis, it transforms these discrete representations back into native speech while retaining the speaker identity. Additionally, we develop an efficient data augmentation method to train the system without demanding a lot of non-native resources. Our system is proved to improve non-native speaker fluency, sound like a native accent, and preserve original speaker identity well.