Pratul P. Srinivasan

Also known as Pratul Srinivasan

Pratul P. Srinivasan is a computer vision and graphics researcher whose work on neural radiance fields (NeRF) and 3D generative models links graphics and machine learning. After a 2014 degree from Duke University, they earned a PhD at UC Berkeley in 2020 under Ren Ng and Ravi Ramamoorthi and coauthored the NeRF paper in 2020 with Ben Mildenhall and Matthew Tancik. Follow-up work included mip-NeRF in 2021, Ref-NeRF in 2022, Zip-NeRF in 2023 and the CAT3D multi-view diffusion system in 2024, and in 2025 they received the SIGGRAPH Significant New Researcher Award and the ACM Grace Murray Hopper Award, both shared with Mildenhall. They are a research scientist at Google DeepMind in San Francisco.

Now at: Google DeepMind
Badges: Bridge

Timeline

2014

Graduating from Duke University

They earn a B.S.E. from Duke University in 2014, working with Sina Farsiu on research problems in medical computer vision.
2020–2021

A Berkeley PhD and two prizes

They receive a PhD from UC Berkeley in 2020, supervised by Ren Ng and Ravi Ramamoorthi, and intern at Google Research with Jon Barron and Noah Snavely during the degree. They share the David J. Sakrison Memorial Prize in 2020 and receive an Honorable Mention for the ACM Doctoral Dissertation Award in 2021.
2020

The NeRF paper

In 2020, they coauthor "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis" with Ben Mildenhall, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi and Ren Ng. The ECCV paper represents a scene with a fully connected deep network that takes a single continuous 5D coordinate as input, and it earns a Best Paper Honorable Mention. NeRF enables novel view synthesis, scene geometry reconstruction and recovery of a scene's reflectance properties.
2021–2022

Mip-NeRF and Ref-NeRF

In 2021, they coauthor mip-NeRF with Jonathan T. Barron, Ben Mildenhall and others; the ICCV paper changes NeRF to output density and radiance for a volume of space, which fixes sampling and aliasing problems, and earns a Best Paper Honorable Mention. In 2022, the CVPR paper Ref-NeRF improves how NeRF represents shiny materials and earns a Best Student Paper Honorable Mention.
2022

Imaging a black hole with NeRF

In 2022, they coauthor "Gravitationally Lensed Black Hole Emission Tomography" with Aviad Levis, Andrew A. Chael, Ren Ng and Katherine L. Bouman. The CVPR paper applies NeRF to the problem of reconstructing the dynamic emissive volume around a black hole.
2023–2024

Zip-NeRF and CAT3D

Zip-NeRF, an ICCV oral and best paper finalist in 2023, combines mip-NeRF 360 with Instant NGP to reconstruct huge scenes. In 2024, they coauthor CAT3D, a NeurIPS oral that uses a multi-view latent diffusion model to generate novel views of a scene and turns them into a 3D representation in as little as one minute.
2025

SIGGRAPH and Hopper awards

In 2025, they and Ben Mildenhall receive the SIGGRAPH Significant New Researcher Award for contributions to new representations for 3D graphics, neural rendering, novel view synthesis and generative models of 3D scenes. The pair also receive the 2025 ACM Grace Murray Hopper Award, which cites contributions to radiance field representations, 3D scene capture and rendering, and neural implicit representations.
As of October 2026

Research scientist at Google DeepMind

They work as a research scientist at Google DeepMind in San Francisco, on NeRF-related research and other problems at the intersection of computer vision, graphics, and machine learning. In 2025, they coauthor Bolt3D, an ICCV paper that generates 3D scenes from one or more input images with a multi-view diffusion model.