Fei-Fei Li created ImageNet, the large visual database that transformed modern computer-vision research.
Li and her research team began the ImageNet project in 2006 while she was at Princeton University. The database organized millions of images according to the WordNet lexical hierarchy, giving researchers a large, consistently labeled resource for training and evaluating visual-recognition systems.
ImageNet became especially influential through the ImageNet Large Scale Visual Recognition Challenge, commonly called ILSVRC. The annual competition began in 2010 and required systems to classify and localize objects across a large set of categories.
The 2012 competition brought a major shift toward deep convolutional neural networks, although the competition itself did not create those networks. ImageNet’s scale, labels, and public benchmark helped establish data-intensive deep learning as a dominant approach in computer vision.