Dialect Matters: Cross-Lingual ASR Transfer for Low-Resource Indic Language Varieties

TL;DR

This paper empirically investigates cross-lingual ASR transfer across Indic dialects, revealing that dialectal similarity influences performance but is not the sole factor, and that fine-tuning on small dialect data can be highly effective.

Contribution

It provides a comprehensive analysis of cross-lingual transfer for low-resource Indic dialects, including a case study on Garhwali and insights into bias and error patterns in ASR models.

Findings

Performance improves with reduced phylogenetic distance but is not solely determined by it.

Fine-tuning on small dialect datasets can match results from larger related language datasets.

Analysis reveals bias toward pre-training languages and common transcription errors.

Abstract

We conduct an empirical study of cross-lingual transfer using spontaneous, noisy, and code-mixed speech across a wide range of Indic dialects and language varieties. Our results indicate that although ASR performance is generally improved with reduced phylogenetic distance between languages, this factor alone does not fully explain performance in dialectal settings. Often, fine-tuning on smaller amounts of dialectal data yields performance comparable to fine-tuning on larger amounts of phylogenetically-related, high-resource standardized languages. We also present a case study on Garhwali, a low-resource Pahari language variety, and evaluate multiple contemporary ASR models. Finally, we analyze transcription errors to examine bias toward pre-training languages, providing additional insight into challenges faced by ASR systems on dialectal and non-standardized speech.