LyricWhiz: Robust Multilingual Zero-shot Lyrics Transcription by Whispering to ChatGPT
This work addresses the challenging and emerging task of multilingual lyrics transcription for researchers and developers, offering a novel training-free method and a new dataset, though it is incremental as it builds on existing models.
The authors tackled the problem of multilingual zero-shot lyrics transcription by combining Whisper for audio transcription and GPT-4 for contextual correction, achieving state-of-the-art performance with significant reductions in Word Error Rate in English and effective transcription across multiple languages.
We introduce LyricWhiz, a robust, multilingual, and zero-shot automatic lyrics transcription method achieving state-of-the-art performance on various lyrics transcription datasets, even in challenging genres such as rock and metal. Our novel, training-free approach utilizes Whisper, a weakly supervised robust speech recognition model, and GPT-4, today's most performant chat-based large language model. In the proposed method, Whisper functions as the "ear" by transcribing the audio, while GPT-4 serves as the "brain," acting as an annotator with a strong performance for contextualized output selection and correction. Our experiments show that LyricWhiz significantly reduces Word Error Rate compared to existing methods in English and can effectively transcribe lyrics across multiple languages. Furthermore, we use LyricWhiz to create the first publicly available, large-scale, multilingual lyrics transcription dataset with a CC-BY-NC-SA copyright license, based on MTG-Jamendo, and offer a human-annotated subset for noise level estimation and evaluation. We anticipate that our proposed method and dataset will advance the development of multilingual lyrics transcription, a challenging and emerging task.