Object detection neural network based on yolo and brightness enhancement network for railway tool detection
The source code is being sorted and will be released later...
The dataset is annotated in COCO format.
The folder .\VOCdevkit\VOC2007\JPEGImages is the dataset image,
the folder .\VOCdevkit\VOC2007\Annotations is the label of the dataset,
and the folder .\VOCdevkit\VOC2007\ImageSets\Main is the division of the corresponding training set, validation set and test set,
so the test set pictures are no longer listed separately.
.\model_data stores the trained model Ryolo.pth of this project, and the detected tool types are stored in classes.txt.