Ben Mildenhall
Co-founder of World Labs and first author of the NeRF paper on neural radiance fields for view synthesis
- Co-founder, working on graphics and 3D computer vision, World Labs12
Ben Mildenhall is one of the co-founders of World Labs, where Mildenhall works on problems in graphics and 3D computer vision. Mildenhall was a research scientist at Google Research from 2021 to 2023, and is the first author of the 2020 NeRF paper, which represents scenes as neural radiance fields for view synthesis.123
Career
Mildenhall received a PhD from UC Berkeley in 2020, advised by Ren Ng and supported by a Hertz fellowship, and had earlier internships in Marc Levoy's group at Google Research in 2017 and at Fyusion in 2018. From 2021 to 2023 Mildenhall was a research scientist at Google Research.2
World Labs names Mildenhall as one of its founders, alongside Fei-Fei Li, Justin Johnson and Christoph Lassner. Mildenhall's own site describes current work on graphics and 3D computer vision at World Labs.12
Research
Mildenhall was first author of the 2020 NeRF paper on neural radiance fields. Later papers with Jonathan Barron and colleagues extended the approach: Mip-NeRF 360 (2021) handled unbounded scenes, and Zip-NeRF (2023) combined anti-aliasing with grid-based representations.356
Mildenhall also co-authored DreamFusion (2022), which generates 3D content from text prompts using a 2D diffusion model, and ReconFusion (2024), which uses diffusion priors for 3D reconstruction from few images.42
Work
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis (external site: arxiv.org)Published research
First author, with Pratul Srinivasan, Matthew Tancik, Jonathan Barron, Ravi Ramamoorthi and Ren Ng, of the 2020 paper that represents a scene as a neural radiance field for synthesizing novel views; Mildenhall's site lists it at ECCV 2020 with a Best Paper Honorable Mention.32
DreamFusion: Text-to-3D using 2D Diffusion (external site: arxiv.org)Published research
Co-author, with Ben Poole, Ajay Jain and Jonathan Barron, of the 2022 paper on generating 3D content from text using a 2D diffusion model.4
Career timeline
2020
Received a PhD from UC Berkeley, advised by Ren Ng.2
Mar 2020
Posted the NeRF paper on neural radiance fields as first author.3
2021
Began work as a research scientist at Google Research.2
Nov 2021
Co-authored Mip-NeRF 360 on unbounded anti-aliased neural radiance fields.5
Sep 2022
Co-authored DreamFusion on text-to-3D generation with 2D diffusion.4
Apr 2023
Co-authored Zip-NeRF on anti-aliased grid-based neural radiance fields.6
Other documented roles
Roles the sources document but that are not confirmed as current.
- Research scientist (2021 to 2023), Google Research2
Honors and recognition
Related people
More sourced profiles at the same organization, in the same field, or from the same school.
- Christoph LassnerAlso at World Labs
- Aditya RameshImage, video, and 3D generation
- Adam GleaveAlso from UC Berkeley
- Fei-Fei LiAlso at World Labs
- Alex NicholImage, video, and 3D generation
- Alex SmolaAlso at Google Research
- Justin JohnsonAlso at World Labs
- Alexei A. EfrosAlso from UC Berkeley
- Ali GhodsiAlso from UC Berkeley
- Abhimanyu DasAlso at Google
- Amir HertzAlso at Google
- Anca DraganAlso from UC Berkeley
Is this you?
Profiles cover public professional work only, and every line cites a source. If something is out of date, missing or wrong, send a correction with a link to an official page and we will update it. Suggest a correction
Add this badge to your website or GitHub profile README. It links back to this page. It is not an endorsement or a rating.
Markdown (for a README)
[](https://unitedstatesofamericasuperintelligence.com/people/ben-mildenhall/)
HTML
<a href="https://unitedstatesofamericasuperintelligence.com/people/ben-mildenhall/"><img src="https://unitedstatesofamericasuperintelligence.com/badges/profile.svg" alt="Ben Mildenhall — Profile on USASI" height="20"></a>
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.