MIT EECS | Morais and Rosenblum Undergraduate Research and Innovation Scholar
Efficient ML for Audio-Visual Speech Recognition
- Computer Architecture
Anantha P. Chandrakasan
Audio-Visual Speech Recognition (AVSR) is a domain of machine learning that utilizes the visual modality to improve automatic speech recognition. My project aims to explore methods make AVSR models for efficient and deployable on resource-constrained devices.
Through this SuperUROP, I hope to apply what I learned in classes such as CV, NLP, TinyML to my project and get a taste of what doing research is like.