smallcorgi
Can't understand the angle trained in network
In data_processing.py, 43-45 lines shows how to get theta than trianed, but i don't understand the equal in file. theat_train should be rotation_ray - angle_obj, could you explain your method?
Hi @sdukaka,
It just orientates +pi/2, and it will subtract pi/2 in the final output.
But this angel is only the degree between camera and object center and what we want to use in training stage is the difference of theta_ray - rotation_y(the obj's own rotation)
Yes, the outputs of network are 3D dimensions and observation angle of the object.
@sdukaka Hi, have you figured out what the angle exactly means? I can't understand it at all. Could you please help me understand what we should train for?
I think the angle what we should train for is between camera and object center.
jaykay [email protected] 于2019年1月23日周三 下午4:46写道：
@sdukaka https://github.com/sdukaka Hi, have you figured out what the angle exactly means? I can't understand it at all. Could you please help me understand what we should train for?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/smallcorgi/3D-Deepbox/issues/7#issuecomment-456717948, or mute the thread https://github.com/notifications/unsubscribe-auth/AIcfQnkVS48Xcv2UobGbQzSC0S30NfS9ks5vGCF2gaJpZM4RNPzb .