SpadeLiu

Why not use grid_sample?

1

I think the function in PyTorch grid_sample can get the value given the position, and that may be more convenient

Error about the “q” in the MultiheadAttentionRelative module

3

Hi, I met a problem when running the codes. Please refer to Line 37 to Line 63 of "module/attention.py". Here you list two conditions: one for self-attention and another for...

Hi, I come from your paper "Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation". I want to generate some sparse disparity labels but some errors occur when I...

About the number of images on KITTI used for domain adaption

2

Thanks for your wonderful works and codes! I just wonder if you use the total 200 training pairs on KITTI 2015 for domain adaption, or split the 200 pairs into...

Could you please analyze why MS-GCNet is much better than MS-PSMNet?

3

As in the experiments, MS-GCNet is much better than MS-PSMNet. But the parameters for cost aggregation of them are close, 1.9M vs 2.3M? What is the reason? Is it because...

About Training Schedule

4

Hi, Thanks for your work. I wonder in the training phase, which interval is used to sample 'n' in algorithm 1. 1 or 20 ? In other words, can 'n'...