Prasanta Kumar Ghosh

This paper investigates the effectiveness of incorporating synthetic speech data in Automatic Speech Recognition (ASR) Systems for three Indic languages by analyzing performance gains, script sources, speech synthesis models, and voice cloning.

The paper introduces a new inclusive, multimodal Hindi ASR benchmark with real-world recordings and diverse demographic groups, enabling more robust and realistic evaluation.

This paper proposes a novel pretraining method for Acoustic-to-Articulatory Inversion (AAI) using Phoneme Labels, Articulatory Feature Labels, and Critical-articulator Labels, improving performance and reducing inference costs in low-resource settings.

This paper proposes a multimodal framework for jointly improving Automatic Speech Recognition (ASR) and Dialect Identification (DID) in Indian languages using a Bottleneck Encoder, RoBERTa encoder, gating mechanism, and attention encoder.