Skip to content

chequanghuy/TwinLiteNet

Repository files navigation

TwinLiteNet: Efficient and Lightweight Model for Drivable Area & Lane Segmentation

🔥NEW🔥 🔴 TwinLiteNetPlus 🔴 has been officially released!

Check it out now at arXiv GitHub for enhanced performance and new features! 🎉🔥

TwinLiteNet is a lightweight and efficient deep learning model designed for Drivable Area Segmentation and Lane Detection in self-driving cars. This repository provides the code and resources needed to train, evaluate, and deploy TwinLiteNet.


🚀 Requirements

Make sure you have the required dependencies installed:

pip install -r requirements.txt

📂 Data Preparation

  1. Download images from BDD100K Dataset.
  2. Download annotations:

Dataset Structure

/data
    bdd100k
        images
            train/
            val/
            test/
        segments
            train/
            val/
        lane
            train/
            val/

🏗️ Pipeline Overview


🔥 Training

Train the model using the command:

python3 train.py

🎯 Testing

Evaluate the model performance using:

python3 val.py

🖼️ Inference

Perform inference on images:

python3 test_image.py

🔍 Visualization

Drivable Area Segmentation

Lane Detection


📜 Acknowledgment

This work is inspired by:


📖 Citation

If you find our work helpful, please consider starring ⭐ this repository and citing our paper:

@INPROCEEDINGS{10288646,
  author={Che, Quang-Huy and Nguyen, Dinh-Phuc and Pham, Minh-Quan and Lam, Duc-Khai},
  booktitle={2023 International Conference on Multimedia Analysis and Pattern Recognition (MAPR)}, 
  title={TwinLiteNet: An Efficient and Lightweight Model for Driveable Area and Lane Segmentation in Self-Driving Cars}, 
  year={2023},
  pages={1-6},
  doi={10.1109/MAPR59823.2023.10288646}
}

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages