Learning Cross-lingual Mappings for Data Augmentation to Improve Low-Resource Speech Recognition
This addresses data scarcity for low-resource languages in speech recognition, though it is incremental as it extends existing mapping techniques to end-to-end systems.
The paper tackles the problem of low-resource speech recognition by learning cross-lingual mappings to transliterate source languages into the target language without parallel data, using this for data augmentation, resulting in up to a 5% relative gain over a baseline monolingual model.
Exploiting cross-lingual resources is an effective way to compensate for data scarcity of low resource languages. Recently, a novel multilingual model fusion technique has been proposed where a model is trained to learn cross-lingual acoustic-phonetic similarities as a mapping function. However, handcrafted lexicons have been used to train hybrid DNN-HMM ASR systems. To remove this dependency, we extend the concept of learnable cross-lingual mappings for end-to-end speech recognition. Furthermore, mapping models are employed to transliterate the source languages to the target language without using parallel data. Finally, the source audio and its transliteration is used for data augmentation to retrain the target language ASR. The results show that any source language ASR model can be used for a low-resource target language recognition followed by proposed mapping model. Furthermore, data augmentation results in a relative gain up to 5% over baseline monolingual model.