Ensure your Python version is 3.9.18 and install the following dependencies:

pip install ultralytics
pip install PyQt6 qt_material opencv-python

It's recommended to install ultralytics first, then uninstall it to use the project's custom code:

pip uninstall ultralytics

Project directory structure:

.
├─model       # Pretrained models
├─UI          # Software UI
└─ultralytics # Model code

Our training dataset is from the Obstacle Dataset, filtered to 7915 images suitable for autonomous driving scenarios. The dataset includes 15 obstacle types such as "stop_sign", "person", "bicycle", "bus", "truck", "car", "motorbike", "reflective_cone", "ashcan", "warning_column", "spherical_roadblock", "pole", "dog", "tricycle", "fire_hydrant", etc.

To improve the robustness of our model, we used Albumentations for jitter augmentation on the bounding boxes in our dataset.

To train the final model, run:

python train.py

To optimize our model for different aspects:

• SlimNeck: Used to make the model lightweight, suitable for vehicle-mounted scenarios.
• GAM: Used to capture more features.
• WIOU: Used to handle inaccuracies in the dataset annotations.

Performance comparison of models:

To run the software UI:

cd UI
python main.py

1. Sohan M, Sai Ram T, Reddy R, et al. A Review on YOLOv8 and Its Advancements[C]//International Conference on Data Intelligence and Cognitive Informatics. Springer, Singapore, 2024: 529-545.
2. Li H, Li J, Wei H, et al. Slim-neck by GSConv: A better design paradigm of detector architectures for autonomous vehicles[J]. arXiv preprint arXiv:2206.02424, 2022.
3. Liu Y, Shao Z, Hoffmann N. Global attention mechanism: Retain information to enhance channel-spatial interactions[J]. arXiv preprint arXiv:2112.05561, 2021.
4. Tong Z, Chen Y, Xu Z, et al. Wise-IoU: bounding box regression loss with dynamic focusing mechanism[J]. arXiv preprint arXiv:2301.10051, 2023.
5. Stein G P, Mano O, Shashua A. Vision-based ACC with a single camera: bounds on range and range rate accuracy[C]//IEEE IV2003 intelligent vehicles symposium. Proceedings (Cat. No. 03TH8683). IEEE, 2003: 120-125.