This repository contains an implementation of the LeNet-5 model in TensorFlow, originally introduced in the 1998 paper Gradient-Based Learning Applied to Document Recognition by Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner.
LeNet-5 is one of the earliest convolutional neural network (CNN) architectures, designed for handwritten digit recognition tasks. The model in this repository has been trained on the MNIST dataset, which is widely used as a benchmark dataset for evaluating machine learning algorithms in digit classification.
- Implementation: reproduction of the original LeNet-5 architecture as described in the paper.
- Pre-Trained Model: Includes a pre-trained LeNet-5 model on the MNIST dataset.
The above image showcases the predictions made by the LeNet-5 model on the MNIST dataset. Each cell represents a test image, along with:
- True Label (True): The actual digit in the image.
- Predicted Label (Pred): The digit predicted by the model.
LeNet-5 is one of the first convolutional neural network (CNN) architectures, introduced in the 1998 paper Gradient-Based Learning Applied to Document Recognition by Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner.