Lucksong

Error:bad argument #1 to 'copy' (sizes do not match

1

I want to compute flow according to github, my code looks like this: ```torch 1 spynet = require('spynet') 2 flowX = require 'flowExtensions' 3 computeFlow = spynet.setup(960, 720, 'sintelFinal') 4...

Hi, thanks for your work, I intend to apply it to my job and I have some questions to confirm. Which image is the optical flow referenced to in the...

feature map size after upsample is 2h or 2(h+1)?

3

As your code 'self.upscore2 = nn.ConvTranspose2d( n_class, n_class, 4, stride=2, bias=False)' for example. According to the deconvolution formula described in the pytorch documentation： Hout=(Hin−1)∗stride[0]−2∗padding[0]+kernel_size[0]+output_padding[0] Hout should be 2*(Hin+1) not 2*Hin

weight[range(in_channels),range(out_channels),:,:,] = filt if in_channels is not equal to out_channels, the code reported an error "IndexError: shape mismatch: indexing arrays could not be broadcast together with shapes (32,) (64,)" if...

Confirm how to use the result flow to warp another image(Pytorch).

1

Hi, thanks for your code. I just want to confirm how the obtained optical flow should be used to image warp. ``` script_pwc.py './data/frame_0010.png' './data/frame_0011.png' './tmp/frame_0010.flo' ``` As the code...

Model list retrieval failed

1

I downloaded graspit! But encountered this problem. In addition, there is no graspit.pro file in my downloaded version of graspit. Can someone solve this problem?

Why there is nn.softmax before nn.crossentroyloss in infogan?

2

In 111 line of the implementaion of infogan, nn.softmax is added to aux_layer. But the returned data of aux_layer is used to compute crossentropy loss in 277 line. Doesn`t nn.crossentropyloss...

The meaning of each parameter of the six-dimensional transformation matrix

2

I have a question about the tform in kitti single model. It is a six-dimensional matrix composed of Euler angles and translation vectors,but what is the specific meaning of each...

Some coordinates will fall outside the image area during forward splat

1

Hi, thanks to your work, it is great while I have some puzzles. During forward splat, we can compute the target coordinates according to the camera intrinsic and extrinsic. After...