Assistant Professor · EECS, York University

Shweta Mahajan

I am a professor in the EECS department at York University. Before this, I was a Machine Learning Researcher at Qualcomm AI Research. I was also a postdoctoral researcher and a Vector Postdoctoral Affiliate in the Vision Group at University of British Columbia advised by Prof. Leonid Sigal and Prof. Kwang Moo Yi.

I obtained my Ph.D. under the supervision of Prof. Stefan Roth, Ph.D. in the Visual Inference Group, Technische Universität Darmstadt.

I received my M.Sc. from Saarland University where I was a part of the Machine Learning Group and the Max Planck Institute of Informatics.

Portrait of Shweta Mahajan

News

Hiring

I am seeking strong, research-dedicated PhD and Master’s (MASc/MSc) students to join my group in the Department of Electrical Engineering and Computer Science at York University. Focus Areas include generative modeling (diffusion, normalizing flows, VAEs), multimodal AI, scene understanding, and physical reasoning. A strong foundation in linear algebra, probability, and optimization, along with strong hands-on proficiency in PyTorch/Python, is required. Prior publication experience at major venues (CVPR, ICCV, ECCV, NeurIPS, ICLR) is highly valued for PhD applicants. If you are interested in working with me, please apply to the York EECS graduate program. Send me an email with your CV, academic transcripts, and a concise summary of your research interest, using the tag [Prospective Student - ] in your email subject.

Research

I am interested in computer vision and machine learning, specifically in deep generative models (diffusion models, normalizing flows, variational methods, GANs) for multimodal representation learning.

Publications

CVPR '24 ·Highlight

Unsupervised Keypoints from Pretrained Diffusion Models

Eric Hedlin, Gopal Sharma, Shweta Mahajan, Xingzhe He, Hossam Isack, Abhishek Kar, Helge Rhodin, Andrea Tagliasacchi, Kwang Moo Yi

One can leverage the semantic knowledge within diffusion models to find keypoints across images of a similar kind.

paper · arxiv · code

Honors & Awards