"All images and labels" combines all the images and their labels in format of yolo format.
"BestModel.pt" is the weights of yolov5 neural network trained based on this dataset(randomly choose 80% of the full dataset). The mAP of this model is around 57%.
"TestResult_7.31" is the detection result of this model in test dataset(randomly choose 20% of the full dataset).
"yolov5 modified" is the source code to generate labeled video and all ".npy" file which includes the coverage data.