Saurabh Saxena
Research @ Google DeepMind
I am a Staff Research Engineer at Google DeepMind, where I lead research at the intersection of computer vision and generative modeling. My work in generative AI explores the potential of diffusion models for creating controllable image and video content (aka world models). On the scene understanding front, my research is centered on methods for recovering intrinsic properties of a scene, such as its geometry and motion, as well as estimating camera parameters. Prior to this, I played a pivotal role in building and launching TensorFlow 2, leading foundational aspects like automatic differentiation, control flow, and eager execution.
News
- June 2025 Our paper on motion segmentation for robust SfM was accepted at ICCV 2025.
Selected Publications
Photorealistic text-to-image diffusion models with deep language understanding
Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, Jonathan Ho, David J Fleet, Mohammad Norouzi
NeurIPS 2022 SPOTLIGHT
Large-scale evolution of image classifiers
Esteban Real, Sherry Moore, Andrew Selle, Saurabh Saxena, Yutaka Leon Suematsu, Jie Tan, Quoc V Le, Alexey Kurakin
ICML 2017