Fei-Fei Li

Overview

Fei-Fei Li is a computer scientist whose work on large-scale visual data changed how machines learn to recognize what is in a picture. She was born in Beijing in 1976, grew up in Chengdu, and moved to the United States as a teenager, living in Parsippany, New Jersey^[2].

Education and early research

Li studied physics at Princeton University, graduating with a B.A. and high honors in 1999, then earned a PhD in electrical engineering at Caltech in 2005 under the supervision of Pietro Perona and Christof Koch^[1]. Training first in physics and then in engineering left her with a measurement-first habit of mind: how do you describe an image, how do you collect enough of them, and how do you know whether a system has actually improved?

ImageNet

Her best-known contribution is ImageNet, a large-scale image dataset, and the ImageNet Challenge that grew alongside it. She began developing ImageNet in 2007, while she was at Princeton^[1]. The dataset was presented in a 2009 CVPR paper, "ImageNet: A Large-Scale Hierarchical Image Database", written with Jia Deng, Wei Dong, Richard Socher, Li-Jia Li and Kai Li, with Li as senior author^[1][3].

The significance lies in what a shared dataset does to a field. Computer vision had long relied on features designed by hand — researchers guessed which measurements of edges, textures or shapes would matter — and on test sets small enough to be run repeatedly by a single group. A large, labelled, public collection gives many groups the same training material, and a public challenge gives them the same scoreboard. Methods can then be compared directly, and errors can be traced to particular kinds of images rather than argued about in the abstract.

Stanford and public institutions

Li has taught at Stanford since 2009, after assistant professorships at the University of Illinois Urbana-Champaign (2005–2006) and Princeton (2007–2009)^[1]. She is the inaugural Sequoia Professor in Stanford's Computer Science Department^[1]. She directed the Stanford Artificial Intelligence Lab from 2013 to 2018 and is a founding co-director of Stanford's Institute for Human-Centered AI (HAI)^[1].

Industry

From 2017 to 2018 she was Vice President at Google and Chief Scientist of AI/ML at Google Cloud^[1]. In 2024 she co-founded World Labs, an AI company focused on spatial intelligence, and is its CEO^[1]. Spatial intelligence refers to the ability to reason about objects, positions and relations in three-dimensional space, rather than only about text.

Teaching, writing and access

She co-founded AI4ALL in 2017^[1]. In 2023 she published the memoir "The Worlds I See: Curiosity, Exploration and Discovery at the Dawn of AI"^[1].

Recognition

Li is an elected member of the National Academy of Engineering, the National Academy of Medicine and the American Academy of Arts and Sciences, and an ACM Fellow; she has published more than 400 scientific articles^[1]. Her awards include the IEEE PAMI Longuet-Higgins Prize (2019), the IEEE PAMI Thomas Huang Memorial Prize (2022), the VinFuture Prize (2024) and the Queen Elizabeth Prize for Engineering (2025)^[2].

Across datasets, laboratories and companies, the thread running through her career is a single question: how a machine can build a dependable picture of the world.