Training our own dataset with circular annotations.
I have a cancer dataset with a center pixel circular dataset along with the radius of the circle. Does your model only take data as 'path x1,y1,x2,y2 class' or can I change the format for my needs?
you have to change the localization layers to output the center pixels.
or you can generate circles instead of rectangles using x1,y1,x2,y2 information to save effort
What is the current version of tensorflow-gpu does this support? Doesn't seem to work with tensorflow-gpu-1.12.1
can you try with 1.12.0? that is the version I used.
would be helpful if you can post error comments.
could you please tell about the tuning parameters ? I am already train the model with my custom data and I want to increase the model accuracy , so please tell me about model tuning parameters,
sir one more , this training goes very slowly and could you tell me about how can I increase the training speed ??