Reduce and Reconstruct: ASR for Low-Resource Phonetic Languages
This addresses the problem of limited data for ASR in phonetic languages, offering an incremental improvement for low-resource scenarios.
The paper tackles low-resource automatic speech recognition (ASR) for phonetic languages by reducing the output alphabet through acoustically similar grapheme sets and reconstructing it with a finite state transducer, achieving up to 7% relative word error rate reduction with only 10 hours of speech data.
This work presents a seemingly simple but effective technique to improve low-resource ASR systems for phonetic languages. By identifying sets of acoustically similar graphemes in these languages, we first reduce the output alphabet of the ASR system using linguistically meaningful reductions and then reconstruct the original alphabet using a standalone module. We demonstrate that this lessens the burden and improves the performance of low-resource end-to-end ASR systems (because only reduced-alphabet predictions are needed) and that it is possible to design a very simple but effective reconstruction module that recovers sequences in the original alphabet from sequences in the reduced alphabet. We present a finite state transducer-based reconstruction module that operates on the 1-best ASR hypothesis in the reduced alphabet. We demonstrate the efficacy of our proposed technique using ASR systems for two Indian languages, Gujarati and Telugu. With access to only 10 hrs of speech data, we obtain relative WER reductions of up to 7% compared to systems that do not use any reduction.