Paper: Multilingual Speech Recognition For Information Retrieval In Indian Context

ACL ID N04-2001
Title Multilingual Speech Recognition For Information Retrieval In Indian Context
Venue Human Language Technologies
Session Student Session
Year 2004
Authors

This paper analyzes various issues in building a HMM based multilingual speech recognizer for Indian languages. The system is originally designed for Hindi and Tamil languages and adapted to incorporate Indian accented Eng- lish. Language-specific characteristics in speech recognition framework are highlighted. The recognizer is embedded in information re- trieval applications and hence several issues like handling spontaneous telephony speech in real-time, integrated language identification for interactive response and automatic graph- eme to phoneme conversion to handle Out Of Vocabulary words are addressed. Experiments to study relative effectiveness of different al- gorithms have been performed and the results are investigated.