Marcus Brubaker
Also known as Marcus A. Brubaker
Marcus Brubaker is an AI researcher whose work spans generative models and machine-learning methods for three-dimensional molecular reconstruction. He earned a computer science PhD at the University of Toronto in 2011, helped build protein reconstruction algorithms at Structura Biotechnology, and led research at Borealis AI from 2018 to 2020. In 2019, he coauthored a review of normalizing flows; in 2024, he coauthored work on text-guided image and scene editing, and in 2026, he coauthored ORBIT++, a paper on benchmarking structure-from-motion methods with 360° video. He joined Google DeepMind's Toronto team in August 2024 and is a research scientist there and an associate professor at York University.
- Now at
- York University and Google DeepMind
Timeline
-
2011
A computer science PhD
He completes a PhD in computer science at the University of Toronto in 2011.
-
2015
3D molecular reconstruction
In 2015, he presents an iterative method for three-dimensional molecular reconstruction that lets one computer do in a day work that had required hundreds of computers and several weeks.
-
By 2017
Structura and protein reconstruction
By 2017, he co-founds Structura Biotechnology with PhD student Ali Punjani. The company develops machine-learning algorithms to reconstruct high-resolution protein structures, and its cryoSPARC software rapidly determines molecular structures.
-
2018–2020
Research director at Borealis AI
From 2018 to 2020, he serves as Research Director of Borealis AI, a machine-learning research lab founded by the Royal Bank of Canada.
-
2019
Normalizing flows for generative models
In 2019, he coauthors a review explaining how normalizing flows model tractable distributions with efficient, exact sampling and density evaluation.
-
2020–2021
Conference reviewer honors
He receives Top Reviewer recognition from ICML in 2020 and Outstanding Reviewer recognition from the British Machine Vision Conference in 2021.
-
2023–2024
Promotion and text-guided editing
In July 2023, he receives tenure and promotion to Associate Professor at York University. A 2024 paper, Watch your steps, explores local image and scene editing by text instructions.
-
Since August 2024
Research scientist at Google DeepMind
In August 2024, he joins Google DeepMind as a research scientist in Toronto and is an associate professor of computer science at York University. At ICCV 2025, he coauthors RoMo, a paper on robust motion segmentation for structure from motion. In 2026, he coauthors ORBIT++, a paper that benchmarks structure-from-motion methods with 360° video.
Career
- Since 2024Google DeepMindResearch scientist
- Since 2023York UniversityAssociate Professor of Computer Science
- 2018–2020Royal Bank of CanadaResearch Director
- From 2017Structura BiotechnologyCo-founder
Education
- 2011University of TorontoPhD
Sources and credits
Photo: We have no published photo of Marcus Brubaker yet. Send us a free-licensed photo.
- Marcus Brubaker - Vector Institute for Artificial Intelligence
- 3-D vision: UTSC computer scientists develop better way to visualize molecules | University of Toronto
- 2017 Research and Innovation Annual Report - Student Experience
- [1908.09257] Normalizing Flows: An Introduction and Review of Current Methods
- About Me - Marcus A. Brubaker
- Marcus Brubaker - Centre for Innovation in Computing at Lassonde
- Google at ICCV 2025
- ORBIT++: Benchmarking SfM in the Wild with 360° Video


