Semantic Segmentation for Self Driving Cars is a project focused on pixel-level image segmentation for autonomous vehicles. The goal is to accurately classify each pixel in an image, assigning it to a specific object class, such as road, vehicle, pedestrian, or obstacle. This repository contains the code, models, and data for training and evaluating the semantic segmentation model.
Semantic segmentation is a critical component of self-driving car systems. It provides a detailed understanding of the road environment, allowing autonomous vehicles to make informed decisions. This project explores state-of-the-art deep learning techniques, including U-Net and ResNet-based architectures, for achieving high-precision semantic segmentation.
Follow these instructions to set up and run the project on your local machine. Please note that this project requires a GPU for efficient training.
- Python 3.10.11
- TensorFlow (2.12.0)
- NumPy
- Matplotlib
- GPU for training (recommended for faster training)
-
Clone the repository:
git clone https://github.com/ArnabKumarRoy02/Semantic-Segmentation-SDC.git cd Semantic-Segmentation-SDC -
Create a virtual environment (preferably using Conda):
conda create -n venv conda activate venv
-
Download the data from Kaggle
Contributions are welcome! If you'd like to improve this project or fix any issues, please open a pull request or create an issue.
This project is licensed under the MIT License.