Profile

Fei-Fei Li was born in Beijing in 1976 and moved to the United States with her family as a teenager. She earned a bachelor’s degree in physics at Princeton University and a PhD in electrical engineering at the California Institute of Technology, working where computer vision meets cognitive neuroscience.

After faculty posts at the University of Illinois Urbana-Champaign and Princeton, she joined Stanford University in 2009, where she is a professor of computer science and formerly directed the Stanford Artificial Intelligence Lab. During a 2017–2018 sabbatical she served as a vice president at Google and chief scientist of AI/ML at Google Cloud. In 2019 she co-founded the Stanford Institute for Human-Centered Artificial Intelligence (HAI) and became its co-director. In 2024 she co-founded the spatial intelligence company World Labs.

Key contributions

Li is best known for ImageNet. In the mid-to-late 2000s, computer vision research was held back by small training datasets. She argued that if machines were to recognise the world as people do, they needed a large labelled dataset covering everyday object categories. Her team used crowdsourcing to build ImageNet, with more than ten million images across over twenty thousand categories, and published it in 2009.

The ImageNet Large Scale Visual Recognition Challenge (ILSVRC) built around it became the common yardstick for vision algorithms. In 2012 Geoffrey Hinton’s group made its AlexNet breakthrough in that very competition, marking the start of the deep learning era.

She has also published widely on visual scene understanding, image captioning and ambient intelligence in healthcare, and helped drive work that connects vision and language.

Notable work

  • The ImageNet dataset and the ImageNet Large Scale Visual Recognition Challenge.
  • Co-founding AI4ALL, a nonprofit that brings young people from diverse backgrounds into AI.
  • Co-founding Stanford HAI to connect AI research with ethics, policy and the social sciences.
  • Her 2023 memoir, The Worlds I See.
  • Founding World Labs, which launched its first commercial product, Marble, in 2025. It generates exportable 3D environments from text, images or video.

Honours

  • Member of the US National Academy of Engineering, the National Academy of Medicine and the American Academy of Arts and Sciences.
  • 2025 Queen Elizabeth Prize for Engineering, shared with Geoffrey Hinton, Yann LeCun, Yoshua Bengio and others.

Influence

ImageNet changed how AI research is done. It showed that the scale and quality of data matter as much as the algorithm, an idea that runs straight through today’s data-hungry large models.

Li is also one of the most prominent advocates of “human-centred AI.” She has repeatedly argued that AI should augment human capabilities and serve society, with attention to diversity and fairness. She has taken part in US national AI policy discussions and pushed for public-sector access to compute and data.

Her recent focus is spatial intelligence: the view that teaching AI to understand and generate three-dimensional worlds is the next major frontier after language. From helping machines see to helping them understand the space they see, her work keeps returning to one question: how can a machine truly know the world?