Normal-Assisted-Stereo

Normal-Assisted-Stereo copied to clipboard

hi Thans for your great job! And I have a easy question. where should I get the test scene. I down the scannet data by [this link ](https://drive.google.com/drive/folders/1PTi37xlPxqhHNyxs_4xiGGj1OsnTQhWD). Howerer, this...

tyjiang1997

Can't reproduce the performance on the ScanNet dataest.

1

Hi, I tried to train the model on the ScanNet dataset, but we got a very worse result, compared with the result reported in the paper. We got 0.2440 on...

lqzhao

Hi! Thank you for share the code. I wonder have you trained or tested the model on kitti? If yes I just want to get a rough sense how well...

ykk136

vectorization

2

Hi, thanks for the amazing work. Do you know how to vectorize the code for computing normal from depth, it is pretty slow using for loop. I tried to vectorize...

qiminchen

error in computing depth_grad_2?

1

Hi, shouldn't `c_x` and `c_y` here be the `u_c` and `v_c` of the intrinsic matrix according to equation 7 in the paper? https://github.com/udaykusupati/Normal-Assisted-Stereo/blob/491c0d3d2efd0e6c9f97cf84b1688266b9c201d6/models/ConsModule.py#L52-L53

qiminchen

Thank you for your contribution!!! As I check out the google drive but nothing found, could you please provide the pretraiend model training of yourself? Thanks!

ruyanyinian

Hi, When I try to download the file [scannet.tar.gz](https://drive.google.com/drive/folders/1PTi37xlPxqhHNyxs_4xiGGj1OsnTQhWD?usp=sharing) it says "This folder is empty". Is it still possible to obtain the datasets? Thanks, Daniel

dlichy

Dear author, Thanks for your work! When I reproduce the results in the paper, I found that link 'https://drive.google.com/drive/folders/1PTi37xlPxqhHNyxs_4xiGGj1OsnTQhWD?usp=sharing' is empty. Could you share permissions on the project? Thanks!! 

Qianqian3764