Hi! I’m Maharaj Brahma, a second-year PhD student at Natural Language and Information Processing (NLIP) Lab in the Department of Computer Science & Engineering at the Indian Institute of Technology Hyderabad (IITH). I’m fortunate to be supervised by Prof. Maunendra Sankar Desarkar and Dr. Anoop Kunchukuttan. My research interests include Multilingual NLP focusing on low-resource languages, Culture Intelligence in Large Language Models (LLMs), and Machine Translation for low-resource languages. I am also interested in building resources for low-resource languages.
Prior to joining the Ph.D. program, I completed my Master of Technology (M.Tech.) at Central Institute of Technology Kokrajhar (CITK), CFTI, Deemed to be University under MoE, India. I was fortunate to be supervised by Prof. Sanjib Narzary, and I worked on Machine Translation for the under-resourced Indian language Bodo. I served as a Teaching Assistant (TA) for the master’s course Advanced Computer Network Lab (PCSE271) and the undergraduate course Programming for Problem Solving Lab (UCSE271). I was a TA for the master’s course Mobile and Pervasive Computing (PCSE115), instructed by Prof. Pranav Kumar Singh.
In 2020, I had the good fortune to co-found a startup “DigitalOma” along with my friends.
I received my Bachelor of Technology (B.Tech.) in Computer Science & Engineering from CIT Kokrajhar, India in 2019 and worked on a thesis titled “English-Bodo Neural Machine Translation using Attention Mechanism”.
News:
- EMNLP 2023. Super thanks to my awesome senior Kaushal and supervisor Prof. Maunendra for their constant guidance and support.My first paper got accepted! Paper titled SelectNoise: Unsupervised Noise Injection to Enable Zero-Shot Machine Translation for Extremely Low-resource Languages accepted at
- Paper Accepted! Paper titled AutoBookFinder: A Case Study of Automated Book Rack Identification in Library through UGV accepted at
- Selected as Virtual Student Volunteer for
- Paper Accepted! Paper titled AI and Blockchain-based Source Code Vulnerabilty Detection and Prevention System for Multiparty Software Development accepted at
- Paper Accepted! Paper titled IntelliStore: IoT and AI-based Intelligent Storage Monitoring for Perishable Food accepted at
- Hugging face and GitHubData Resource! Open source Bodo Words Corpus - Now available at
- Selected for
- Selected for
- Presented poster titled A Computational Approach for the Tonal Identification in Bodo Language alongside
- Paper Accepted! Paper titled GROUP: Global RObUst and Privacy Preserved Model for Diabetes Prediction accepted at
- Paper Accepted! Our work Generating Monolingual Dataset for Low Resource Language Bodo from old books using Google Keep accepted at