Robust Speech Recognition via Large-Scale Weak Supervision
Radford et al., 2023, Robust Speech Recognition via Large-Scale Weak Supervision (Proceedings of the 40th International Conference on Machine Learning (ICML 2023), PMLR 202)
Subscriber deck — free during early access
1 cards in “3.4 Multi-lingual Speech Recognition” · whole deck
3.4 Multi-lingual Speech Recognition
- How does Whisper’s error rate in a language depend on training data?Log WER is strongly correlated with log training hours for that language (\(R^2 = 0.83\)); WER roughly halves for every 16× more data. The worst outliers were languages with unique scripts and distant from the Indo-European majority, such as Hebrew, Telugu, Chinese and Korean.
- Radford et al., 2023, Robust Speech Recognition via Large-Scale Weak Supervision (Proceedings of the 40th International Conference on Machine Learning (ICML 2023), PMLR 202), §3.4
Source
- Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey and Ilya Sutskever. 2023. Robust Speech Recognition via Large-Scale Weak Supervision. In Proceedings of the 40th International Conference on Machine Learning (ICML 2023), PMLR 202, pp. 28492–28518. https://proceedings.mlr.press/v202/radford23a.html