LeNet
lastmod 2026-11-10created 2015-12-302 edits
LeNet is an early convolutional neural network designed by Yann LeCun for handwritten digit recognition. Its architecture alternates convolutional layers and subsampling (pooling) layers and ends with fully connected layers; LeNet-5 is the representative version, and together with the MNIST dataset it serves as a textbook example of CNNs. The basic structure of later deep CNNs such as AlexNet and GoogLeNet originated here.