hi , after I dived into the detect_face.py, I found this line is unnecessary:
orgimg = cv2.cvtColor(orgimg, cv2.COLOR_BGR2RGB)
then the confience is increased than before , ie. more faces are detetcted correctly.
so this file is a bit more confused/complex than official yolov5 project's one detect.py
hi , after I dived into the detect_face.py, I found this line is unnecessary:
then the confience is increased than before , ie. more faces are detetcted correctly. so this file is a bit more confused/complex than official yolov5 project's one detect.py