Traffic Sign Detection

Data : https://www.kaggle.com/datasets/ahemateja19bec1025/traffic-sign-dataset-classification

Pre-trained Models used:

• VGG16
• Xception
• ResNet50

Data Preparation:

1. Loaded the traffic sign images from the directory structure.
2. Used ImageDataGenerator to rescale the images and split the data into training and validation sets.

VGG16 Model:

1. Loaded the pre-trained VGG16 model without the top layer.
2. Added a custom dense layer for classification.
3. Compiled and trained the model.
4. Saved the model and evaluated its performance, achieving high accuracy.

Xception Model:

1. Loaded the pre-trained Xception model.
2. Added custom layers for classification.
3. Compiled and trained the model with early stopping.
4. Evaluated the model, achieving reasonable accuracy but lower than VGG16.

ResNet50 Model:

1. Loaded the pre-trained ResNet50 model.
2. Added custom layers for classification.
3. Compiled the model and trained the model.
4. Evaluated the model, achieving reasonable accuracy but lower than VGG16.