I am a PhD candidate in the MIT EECS department and a research assistant to Dr James Glass in the Spoken Language Systems group at MIT CSAIL. Prior to joining MIT, I was pursuing MSc in Artificial Intelligence at the University of Edinburgh, where I completed my MSc thesis on Vector Space Modeling of Natural Language under the supervision of Dr Shay Cohen.

My current research interests include unsupervised speech processing, automatic speech recognition and language identification.

Infants learn to segment fluent speech into discrete units such as words at the early stages of native language acquisition without much supervision. I am interested in building machines that can acquire language in a completely unsupervised manner. To that end, i draw inspiration from psycholinguistics, cognitive science and machine learning research to build practical probabilistic models of language acquisition.

I also collaborate with Ahmed Ali and Dr Stephan Vogel from Qatar Computing Research Institute on developing large scale ASR and dialect identification systems for Arabic language.



Unsupervised Speech Processing

All humans process vast quantities of unannotated speech and manage to learn phonetic inventories, word boundaries, etc., and can use these abilities to acquire new word. Why can't ASR technology have similar capabilities? Our goal in this research project is to build speech technology using unannotated speech corpora.


Automatic Speech Recognition

Automatic speech recognition (ASR) has been a grand challenge machine learning problem for decades. Our ongoing research in this area examines the use of deep learning models for distant and noisy recording conditions, multilingual, and low-resource scenarios.