Mehar Khurana

I'm currently a research intern at the Toronto Computation Imaging Group (TCIG), where I'm working with Prof. David Lindell and Anagh Malik on curating a single-photon dataset. I am also working with Prof. Saket Anand, Depanshu Sani and Sharat Agarwal on modelling uncertainties in Multi-Object Tracking.

Previously, I have worked with Prof. Deva Ramanan, Prof. James Hays and Neehar Peri on cross-modal pre-training for 3D object detection.

Email  /  CV  /  Scholar  /  GitHub

Research

I am interested in the fields of computer vision, computational imaging, deep learning, and generative AI. My research primarily focuses on understanding and interpreting the physical world through images, using techniques such as object detection and tracking.

Pre-training on pseudolabels generated by 2D foundation models significantly improves 3D object detection accuracy with limited labeled data for both LiDAR-only and multi-modal detectors.

Source code stolen from Jon Barron's website.