You can create your own lane segmentation model using this dataset https://www.kaggle.com/datasets/mehmetokuyar/traffic-sign-dataset , then you can set up a traffic sign detection algorithm by referencing this code.
Note: You can easily use this code in your own projects by changing the class names in the code.
The input parameters can be changed using the command line :
yolo_aug.py -i <input_dir> -w <weight_path> -o <output_dir>
For running image test example :
python3 yolov7_tensorrt_test.py -i path/image.jpg -w path/yolov7.trt -o path/
For running video test example :
python3 yolov7_tensorrt_test.py -v path/video.mp4 -w path/yolov7.trt -o path/