About This Project
This project implements a deep learning model to classify traffic signs using a fine-tuned VGG-16 convolutional neural network (CNN). Built with TensorFlow and Keras, it processes a dataset of traffic sign images, trains a model to recognize 43 distinct classes, and evaluates its performance. The code is provided as a single Python script (traffic_sign_classifier.py) that handles data downloading, preprocessing, training, evaluation, and visualization.
Key Features
- Download and preprocess a traffic sign dataset automatically.
- Fine-tune a pre-trained VGG-16 model for traffic sign classification.
- Train the model and evaluate its accuracy on validation and test datasets.
- Generate visualizations of training progress and sample predictions.
