Recent advances in speech recognition and translation rely on hundreds of thousands of hours of Internet speech data. We argue that state-of-the art accuracy can be reached without relying on web-scale data. Canary - multilingual ASR and speech translation model, outperforms current state-of-the-art models - Whisper, OWSM, and Seamless-M4T on English, French, Spanish, and German languages, while being trained on an order of magnitude less data than these models. Three key factors enables such data-efficient model: (1) a FastConformer-based attention encoder-decoder architecture (2) training on synthetic data generated with machine translation and (3) advanced training techniques: data-balancing, dynamic data blending, dynamic bucketing and noise-robust fine-tuning. The model, weights, and training code will be open-sourced.
@article{arxiv.2406.19674,
title = {Less is More: Accurate Speech Recognition & Translation without Web-Scale Data},
author = {Krishna C. Puvvada and Piotr Żelasko and He Huang and Oleksii Hrinchuk and Nithin Rao Koluguri and Kunal Dhawan and Somshubra Majumdar and Elena Rastorgueva and Zhehuai Chen and Vitaly Lavrukhin and Jagadeesh Balam and Boris Ginsburg},
journal= {arXiv preprint arXiv:2406.19674},
year = {2024}
}