https://github.com/savinrazvan/traffic
This project aims to develop a neural network using TensorFlow to classify traffic signs from images, utilizing the German Traffic Sign Recognition Benchmark (GTSRB) dataset.
https://github.com/savinrazvan/traffic
ai cnn data-augmentation data-preprocessing deep-learning gtsrb image-recognition machine-learning model-evaluation model-training opencv tensorflow traffic-sign-classification
Last synced: 3 months ago
JSON representation
This project aims to develop a neural network using TensorFlow to classify traffic signs from images, utilizing the German Traffic Sign Recognition Benchmark (GTSRB) dataset.
- Host: GitHub
- URL: https://github.com/savinrazvan/traffic
- Owner: SavinRazvan
- License: mit
- Created: 2024-07-30T13:23:44.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2024-07-30T17:13:55.000Z (10 months ago)
- Last Synced: 2025-01-10T02:32:49.024Z (5 months ago)
- Topics: ai, cnn, data-augmentation, data-preprocessing, deep-learning, gtsrb, image-recognition, machine-learning, model-evaluation, model-training, opencv, tensorflow, traffic-sign-classification
- Language: Jupyter Notebook
- Homepage:
- Size: 186 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
### Traffic Sign Classification with TensorFlow
This project aims to develop a neural network using TensorFlow to classify traffic signs from images, utilizing the German Traffic Sign Recognition Benchmark (GTSRB) dataset.
#### Features
1. **Data Preparation**:
- Images are preprocessed using OpenCV for resizing and normalization.
- Data augmentation techniques are applied to enhance the diversity of the training dataset.
2. **Model Development**:
- A convolutional neural network (CNN) is constructed with TensorFlow.
- The architecture includes convolutional layers for feature extraction, pooling layers for downsampling, and fully connected layers for final classification.
3. **Evaluation**:
- The model's performance is validated using a separate test dataset.
- Predictions are made on unseen data to evaluate real-world applicability.
4. **Documentation**:
- Comprehensive documentation of the experimentation process, including hyperparameter tuning and model iterations.
#### Dataset
The GTSRB dataset can be accessed via the following links:
- [GTSRB - Training and Testing Dataset](https://cdn.cs50.net/ai/2023/x/projects/5/gtsrb.zip)
- [GTSRB Small - Training and Testing Dataset](https://cdn.cs50.net/ai/2023/x/projects/5/gtsrb-small.zip)
#### Code Modifications
This implementation includes several enhancements for improved model performance and additional functionality.
#### Additional Resources
For further information, visit the [project page](https://cs50.harvard.edu/ai/2020/projects/5/traffic/).