STRATA: Word Boundaries & Phoneme Recognition From Continuous Urdu Speech using Transfer Learning, Attention, & Data Augmentation
It addresses the expensive and tedious need for hand-labeled phonetic transcriptions in NLP for low-resource languages, offering a supervised solution with incremental improvements.
The paper tackles phoneme recognition for low-resource languages like Urdu by proposing STRATA, a seq2seq framework using transfer learning, attention, and data augmentation, achieving a Phoneme Error Rate of 16.5% and improving state-of-the-art by 1.1% on TIMIT and 11.5% on CSaLT datasets.
Phoneme recognition is a largely unsolved problem in NLP, especially for low-resource languages like Urdu. The systems that try to extract the phonemes from audio speech require hand-labeled phonetic transcriptions. This requires expert linguists to annotate speech data with its relevant phonetic representation which is both an expensive and a tedious task. In this paper, we propose STRATA, a framework for supervised phoneme recognition that overcomes the data scarcity issue for low resource languages using a seq2seq neural architecture integrated with transfer learning, attention mechanism, and data augmentation. STRATA employs transfer learning to reduce the network loss in half. It uses attention mechanism for word boundaries and frame alignment detection which further reduces the network loss by 4% and is able to identify the word boundaries with 92.2% accuracy. STRATA uses various data augmentation techniques to further reduce the loss by 1.5% and is more robust towards new signals both in terms of generalization and accuracy. STRATA is able to achieve a Phoneme Error Rate of 16.5% and improves upon the state of the art by 1.1% for TIMIT dataset (English) and 11.5% for CSaLT dataset (Urdu).