Yinyu Nie

Hi, Its strange. The code is developed under Ubuntu but I think it is the same for Centos. Can you post any debug info?

Hi, We inherit the layout estimation network (LEN) and corresponding loss from CooP paper. If you only train LEN and remove other modules, our network is exactly the LEN in...

Hi, In the paper, we actually ask for camera intrinsics (i.e., cam_K), otherwise, this problem would be extremely ambiguous. For the layout estimation in our demo, our prediction is [here](https://github.com/yinyunie/Total3DUnderstanding/blob/master/demo/outputs/1/recon.png)....

Hi, For the ODN and LEN training, SUN-RGBD is of a small-scale dataset (~5000 images for training), it indeed overfits with training too many epochs, but we did not see...

Hi, I used the Chamfer distance library from [link](https://github.com/chrdiller/pyTorchChamferDistance), where you may find some solutions. Best, Yinyu

The code is developed under Ubuntu, I tested it several times. For a higher version of PyTorch like 1.8, It still can run. So as my colleagues. You can post...

Hi, I think you should follow our paper to train it by stages, just as TMNet referred in our work. A general strategy is to first train the AtlasNet (by...

Hi, Boundary loss will only work for points on open boundaries, which works at the second stage (tmn_subnetworks =2). So it will be 0s if tmn_subnetworks =1. The first stage...

Sorry for the delay. I will update it today. Please wait for a while and check.

Hi, I have updated the pretrained links. Please check again. Will update the preprocessed data links soon as I am considering transferring them to a new drive. In the meantime,...