Anusha Prakash

Background

About Anusha Prakash

I am a PhD Scholar (Electrical Engineering dept) at the Indian Institute of Technology (IIT) Madras, India. My research work is on developing good quality text-to-speech (TTS) synthesisers for Indian languages, especially for low-resource languages in a multilingual manner. I have experience in working with different TTS frameworks, from traditional techniques (unit selection synthesis, HMM-based, conventional neural network-based) to state-of-the-art end-to-end speech synthesisers. My work focuses on incorporating linguistic knowledge and signal processing techniques seamlessly with deep learning techniques to improve synthesis quality and intelligibility. I have also worked on improving the quality of dysarthric speech. I have worked on various government projects. Most recently, I was a Principal Project Officer in a project called "Speech Technologies for Indian languages", which is a part of "National language translation mission (NLTM)-Bhashini". The main goal of this project is to translate technical lectures in English to various Indian languages. I have trained TTS systems for 15 Indian languages (Assamese, Bengali, Bodo, Gujarati, Hindi, Kannada, Malayalam, Manipuri, Marathi, Odia, Punjabi, Rajasthani, Tamil, Telugu, Urdu) and Indian English of different accents. A few of the models have been deployed as part of the project. These models have been integrated with other applications, such as speech-to-speech and video-to-video transcreations.