Fei-Fei Li
todayComputer-vision pioneer who built ImageNet — the benchmark that triggered the deep-learning revolution. Co-director of Stanford's Human-Centered AI Institute, founder of the AI4ALL nonprofit, and a leading voice for grounding AI policy in human-centred values.
- ImageNet (2009) — 14M labeled images that catalyzed modern computer vision.
- Stanford HAI — human-centered AI research and policy.
FAQ
What is ImageNet?
A dataset Fei-Fei Li and her team built between 2007 and 2009 containing roughly 14 million images hand-labelled across 20,000 categories. The annual ImageNet competition became the de facto benchmark for computer vision, and AlexNet's 2012 win on it triggered the deep-learning revolution.
What is Stanford HAI?
The Stanford Institute for Human-Centered AI, co-directed by Li since its 2019 founding. HAI's research and policy work focus on aligning AI with human values, broadening access to AI talent and compute, and informing US AI legislation.
What is AI4ALL?
A nonprofit Li co-founded in 2017 to make AI education accessible to high-school students from groups historically excluded from computing. It runs free summer programmes at universities including Stanford, Princeton, and CMU.
What is human-centered AI?
Li's framing of an AI research and policy agenda that puts human well-being, dignity, and agency at the centre. In practice it means evaluating systems against social-impact criteria, not only accuracy benchmarks, and investing in mixed-discipline teams that include ethicists and domain experts.