Sarah

Why the occupied memory is larger than the` mem = optnet.countUsedMemory(net)` mem.total_size = 4400 M while I using `nvidia-smi` checking the gpu memory occupied up to almost 11000 M

Is the '''image-level feature''' implemented in the wrong way? It should be a global feature with pooling and unpooling

parameter align_corners for nn.Upsample function

7

When I use the Pytorch 0.4.0 to train the model, I got some warnings: ``` UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True...

How to calculate the initial disparity map?

1

Thank you for your wonderful work and codes. Could you help me understand the code below? https://github.com/LongguangWang/PAM/blob/7200dd201047ceb2d98eeea8215b773d11fff8d4/PASMnet/models/modules.py#L199 Why not directly use the sum of all disparity candidates weighted by the...

How about the result of the implementation?

13

Could you tell me that can this implementation achieves the result authors shown in GC-NET paper?

min() function with vol:cuda()

1

after updating cudnn (v4 to v6), I found that the result of min() changed a lot in x range (parameter: -sm_terminate cnn).

The requirement for cpu memory has not indicated

8

Could you please add the instrument for that?