Home

Xavier Thomas (Rohan)

avatar

👋 Hi! I’m currently a Grad Student at Boston University pursuing my Masters in AI. I did my undergrad from Manipal Institute of Technology, India, in Electronics and Instrumentation with a minor in Computational Intelligence, and had developed a keen interest in all things ML during my first year and was fortunate to gain research experience along the way. Prior to joining BU, I worked with the Content and User Understanding team at ShareChat, and was fortunate to work on projects with the Serre Lab (Brown University), Human Dynamics Group (MIT Media Lab, Massachusetts Institute of Technology), ETS, Montreal and FOR.ai (now Cohere for AI).

At BU, I am fortunate to be advised by Prof. Deepti Ghadiyaram and am working on my MS Thesis focused on Vision-Language Models.

🔍 Currently Seeking: Full-time opportunities in Machine Learning/Computer Vision Research/Engineering roles for Summer 2025.

Curriculum Vitae / Email Me

Education

/img/my_imgs/bu

M.S. in Artificial Intelligence
Boston University
2023 - Present

/img/my_imgs/Manipal

B.Tech. in Electronics and Instrumentation (Minor: Computational Intelligence)
Manipal Institute of Technology
2018 - 2022


Publications, Preprints, and Working Papers

(also see Google Scholar)

/img/my_imgs/MAVIC

MAViC: Multimodal Active Learning for Video Captioning
Gyanendra Das, Xavier Thomas, Anant Raj, Vikram Gupta
Preprint, 2022
Paper

/img/my_imgs/Fig_serrelab_resized

Diversity vs. Recognizability: Human-like generalization in one-shot generative models
Victor Boutin, Lakshya Singhal, Xavier Thomas, Thomas Serre
Neural Information Processing Systems (NeurIPS), 2022
Code | Paper

/img/my_imgs/main_cvpr22_new

Adaptive Methods for Aggregated Domain Generalization
Xavier Thomas, Dhruv Mahajan, Alex Pentland, Abhimanyu Dubey
Preprint, 2021
Code | Paper


Experience

Massachusetts Institute of Technology, Research Collaborator

  • Working with the Camera Culture Group at the MIT Media Lab on Decentralized AI.
Jun 2024 - Present

Boston University, Research Assistant

  • Developed edubotics-core - an open-source Python library for building LLM-based chatbots efficiently, focusing on vector storage, retrieval, and processing.
  • Conducting my MS Thesis (advised by Prof. Deepti Ghadiyaram) focused on Vision-Language Models.
  • Code | PyPI
Jun 2024 - Present

Boston University, Teaching Assistant

Jan 2024 - May 2024

ShareChat, Machine Learning Engineer Intern

  • Implemented a Computer Vision feature pipeline that enhanced downstream tasks like classification and content moderation.
  • Developed web applications for AI-generated content creation and editing on ShareChat’s platforms.
  • Created “MAViC,” a Multimodal Active Learning algorithm, optimizing video captioning model data collection and reducing annotation costs.
  • Advisors: Vikram Gupta, Dr. Hisham Cholakkal, Dr. Anant Raj
  • Code | Paper
July 2022 - June 2023

Serre Lab (Brown University), Research Intern

  • Developed a framework to evaluate one-shot generative models against human performance on the Omniglot dataset.
  • Conducted systematic evaluations of generative models using this framework.
  • Advisors: Dr. Thomas Serre, Dr. Victor Boutin
  • Code | Paper
Sept 2021 - May 2022

École de technologie supérieure (ÉTS), Montréal, Research Intern

  • Summer research intern, funded by the Mitacs Globalink Research Internship Program.
  • Worked on Weakly Supervised Semantic Segmentation.
  • Advisor: Dr. Jose Dolz
  • Code
July 2021 - Sept 2021

Massachusetts Institute of Technology, Research Assistant

  • Developed an algorithm to recover domain information in an unsupervised manner, improving performance on domain generalization datasets.
  • Achieved performance gains in datasets like PACS, VLCS, OfficeHome, DomainNet, and TerraIncognita compared to ERM.
  • Advisors: Dr. Abhimanyu Dubey, Dr. Alex Pentland
  • Code | Paper
Jan 2021 - Nov 2022

For.ai (now Cohere For AI), Research Member

  • FOR.ai is a team of scientists and engineers focused on publishing impactful ML research. Our collaborators include researchers from institutions like Google Brain, University of Oxford, and Vector Institute.
  • Worked on an Out-of-Distribution Detection Benchmarking Project, integrating the Glow architecture for a large-scale study of generative models.
  • Advisor: Sheldon Huang
Oct 2020 - Aug 2021

Miscellaneous