Consisaug: A Consistency-based Augmentation for Polyp Detection in Endoscopy Image Analysis (MIML2023)
The code is changed from yolov5, and the detection for the data format is yolov5. Download pretrained yolov5 checkpoint from yolov5 responsitory and move them to ./trained_model. Change the data configuration file coco.yaml. The hyper-parameter file is hyp.scratch-high.yaml.
python train.py
python val.py
python detect.py