vcg-uvic / learn-orientation Goto Github PK

The paper you write is very good, I try to configure the environment to achieve it, but I do not know how the code is running. This problem has troubled me for a long time. I am a new bird in this field. and I want to study it . can you tell me the details of the operation. Thank you！Thank you！Thank you！beg beg
beg

I have been

I'm currently using this code to check the performance under our custom datasets. However, I cannot find
Can I require you to provide an example config file and a kpts file?

According to paper, "We therefore use a numerical approximation of the gradients: when we form the training data we also compute the descriptors for many possible orientations, every 5 degrees in our current implementation. We can then efficiently compute the derivatives in @g/@Theta by numerical differentiation."
what dose "numerical differentiation" means in this step?
how to do this in code exactly if already get descriptor every 5 degree?
Is it possible to implement in tensorflow?