Siddharth Somasundaram

I am a graduate student at MIT Media Lab in Cambridge, MA, advised by Prof. Ramesh Raskar in the Camera Culture group. My research interests are in computational imaging, computer vision, time-of-flight imaging, and inverse problems.

I previously graduated from UCLA Electrical Engineering in 2021, where I worked with Prof. Diana Huffaker on semiconductor nanowire photodetectors and with Prof. Achuta Kadambi on computational camera design.

Email  /  CV  /  Google Scholar  /  Twitter  /  Github

profile photo
Research

I'm interested in understanding how the space-time propagation of light allows us to infer properties of the world around us. Selected papers are highlighted.

PlatoNeRF: 3D Reconstruction in Plato's Cave via Single-View Two-Bounce Lidar
Tzofi Klinghoffer, Xiaoyu Xiang*, Siddharth Somasundaram*, Yuchen Fan, Christian Richardt, Ramesh Raskar, Rakesh Ranjan
CVPR 2024   (Oral ~0.8% Acceptance)
[project page] [video] [code] [data] [paper]

Reconstructing scene geometry from a single view using two-bounce signals captured by a single-photon lidar.

Role of Transients in Two-Bounce Non-Line-of-Sight Imaging
Siddharth Somasundaram, Akshat Dave, Connor Henley, Ashok Veeraraghavan, Ramesh Raskar
CVPR 2023
[project page] [video] [code] [paper]

Show how transients can reduce the number of measurements needed for two-bounce non-line-of-sight imaging.

Detection and Mapping of Specular Surfaces using Multibounce LiDAR Returns
Connor Henley, Siddharth Somasundaram, Joseph Hollmann, Ramesh Raskar
Optics Express 2023
[project page] [code] [paper]

Exploiting the geometry of two- and three-bounce lidar signals to image mirror-like surfaces.

Physics vs. Learned Priors: Rethinking Camera Design for Task-Specific Camera and Algorithm Design
Tzofi Klinghoffer*, Siddharth Somasundaram*, Kushagra Tiwary*, Ramesh Raskar
ICCP 2022
[video] [paper]

Review perspective on the convergence of computational imaging, end-to-end optimization, and physics-based machine learning for imaging system design.


The source code for this website was adapted from the website of Dr. Jon Barron.