#Learning to See by Moving Code for Learning to See by Moving pdf, ICCV 2015.
WARNING: This release has a lot of things hard coded. I will soon make some fixes to improve usability of the code. Please bear with me in the meanwhile. Feel free to e-mail me for any clarifications.
All the parameters on how training/test data is constucted are specified in
import kitti_utils as ku prms = ku.get_prms()
Form the training and test window files
import kitti_new as kn kn.make_window_file(prms)
Once you have these files, you can use the caffe prototxts provided here to train some models.