LeNet was an early convolutional neural network proposed by Lecun et al in the paper Gradient-Based Learning Applied to Document Recognition.

LeNet was designed for handwriting recognition. Many modern convolutional neural network architectures are inspired by LeNet.