Google’s SpecAugment achieves state-of-the-art speech recognition without a language model

Google AI researchers are using visual analysis data augmentation to achieve state-of-the-art speech recognition performance without a language model.Read More