You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The best solution is to train the neural network yolov3 with the objects that the captcha performs. I don't have a good connection or a good computer for training. My country is banned from many internet services such as the training dataset I want to use. I leave you the list of objects, the dataset and some tutorials to train:
Dataset: https://storage.googleapis.com/openimages/web/download.html
Video Tutorial: https://www.youtube.com/playlist?list=PLZBN9cDu0MSk4IFFnTOIDihvhnHWhAa8W
Object classes: bicycle, bridge, bus, car, chimneys, crosswalk, fire hydrant, motorcycle, palm trees, parking meters, stair, taxis, tractors, traffic light, trees
SAlu2s
Hello
YOLOv3 not detect some image , any solution if not detect image object ?
Thank
The text was updated successfully, but these errors were encountered: