edit build.py, and install successfully.

README.md

@@ -86,7 +86,7 @@ We've included caffe pre-trained models. Should you use these pre-trained weight
 [![Predicted flows on MPI-Sintel](./image.png)](https://www.youtube.com/watch?v=HtBmabY8aeU "Predicted flows on MPI-Sintel")
 
 ## Reference 
-If you find this implementation useful in your work, please acknowledge it appropriately and cite the paper using:
+If you find this implementation useful in your work, please acknowledge it appropriately and cite the paper:
 ````
 @InProceedings{IMKDB17,
   author       = "E. Ilg and N. Mayer and T. Saikia and M. Keuper and A. Dosovitskiy and T. Brox",
@@ -97,6 +97,19 @@ If you find this implementation useful in your work, please acknowledge it appro
   url          = "http://lmb.informatik.uni-freiburg.de//Publications/2017/IMKDB17"
 }
 ````
+```
+@misc{flownet2-pytorch,
+  author = {Fitsum Reda and Robert Pottorff and Jon Barker and Bryan Catanzaro},
+  title = {flownet2-pytorch: Pytorch implementation of FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks},
+  year = {2017},
+  publisher = {GitHub},
+  journal = {GitHub repository},
+  howpublished = {\url{https://github.com/NVIDIA/flownet2-pytorch}}
+}
+```
+## Related Optical Flow Work from Nvidia 
+Code (in Caffe and Pytorch): [PWC-Net](https://github.com/NVlabs/PWC-Net) <br />
+Paper : [PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume](https://arxiv.org/abs/1709.02371). 
 
 ## Acknowledgments
 Parts of this code were derived, as noted in the code, from [ClementPinard/FlowNetPytorch](https://github.com/ClementPinard/FlowNetPytorch).

models.py

@@ -107,23 +107,27 @@ def init_deconv_bilinear(self, weight):
         return 
 
     def forward(self, inputs):
+        # input format: 5d tensor, (batch, rgb, img #, x, y)
+        # normalizing imgs
         rgb_mean = inputs.contiguous().view(inputs.size()[:2]+(-1,)).mean(dim=-1).view(inputs.size()[:2] + (1,1,1,))
-
         x = (inputs - rgb_mean) / self.rgb_max
+
+        # squish the img # dimension by concatenating the rgb channels
         x1 = x[:,:,0,:,:]
         x2 = x[:,:,1,:,:]
         x = torch.cat((x1,x2), dim = 1)
 
         # flownetc
-        flownetc_flow2 = self.flownetc(x)[0]
+        flownetc_flow2 = self.flownetc(x)[0]  # only return flow2, since the network returns flow6 to flow2
         flownetc_flow = self.upsample1(flownetc_flow2*self.div_flow)
 
-        # warp img1 to img0; magnitude of diff between img0 and and warped_img1, 
+        # warp img1 to img0; magnitude of diff between img0 and and warped_img1, i.e. brightness error
+        # combine the flow with img1 to 'reconstruct' img0
         resampled_img1 = self.resample1(x[:,3:,:,:], flownetc_flow)
-        diff_img0 = x[:,:3,:,:] - resampled_img1 
+        diff_img0 = x[:,:3,:,:] - resampled_img1  # brightness error
         norm_diff_img0 = self.channelnorm(diff_img0)
 
-        # concat img0, img1, img1->img0, flow, diff-mag ; 
+        # concat img0, img1, warped img1, flow, normalized brightness error
         concat1 = torch.cat((x, resampled_img1, flownetc_flow/self.div_flow, norm_diff_img0), dim=1)
 
         # flownets1

networks/channelnorm_package/build.py

@@ -24,8 +24,9 @@
     package=False,
     relative_to=this_folder,
     define_macros=Defines,
-    extra_objects=[os.path.join(this_folder, Object) for Object in Objects]
+    extra_objects=[os.path.join(this_folder, Object) for Object in Objects],
+    extra_compile_args=['-std=c99']
 )
 
 if __name__ == '__main__':
-    ffi.build()
+    ffi.build()

networks/channelnorm_package/functions/channelnorm.py

@@ -12,7 +12,8 @@ def forward(ctx, input1, norm_deg=2):
 
         channelnorm.ChannelNorm_cuda_forward(input1, output, norm_deg)
         ctx.save_for_backward(input1, output)
-
+        ctx.norm_deg = norm_deg
+
         return output
 
     @staticmethod

networks/correlation_package/build.py

@@ -24,7 +24,8 @@
     package=False,
     relative_to=this_folder,
     define_macros=Defines,
-    extra_objects=[os.path.join(this_folder, Object) for Object in Objects]
+    extra_objects=[os.path.join(this_folder, Object) for Object in Objects],
+    extra_compile_args=['-std=c99']
 )
 
 if __name__ == '__main__':

networks/resample2d_package/build.py

@@ -24,8 +24,9 @@
     package=False,
     relative_to=this_folder,
     define_macros=Defines,
-    extra_objects=[os.path.join(this_folder, Object) for Object in Objects]
+    extra_objects=[os.path.join(this_folder, Object) for Object in Objects],
+    extra_compile_args=['-std=c99']
 )
 
 if __name__ == '__main__':
-    ffi.build()
+    ffi.build()

requirements.txt

@@ -0,0 +1,3 @@
+setproctitle==1.1.10
+tensorboardX==1.2
+colorama==0.3.9