tsunghan-wu / depth-completion Goto Github PK

Are you using a temporal sequence as the input of the network?

File "/Depth-Completion/depth_completion/data/data_loader.py", line 43, in _load_data_name_matterport
for x in os.listdir(self.data_root):
FileNotFoundError: [Errno 2] No such file or directory: '/work/kaikai4n/new_matterport/v1

hi teacher wu:
Could you please provide a data link to Google drive or Baidu Cloud, something wrong when i use ladder

I have got the access to the original dataset, could you kindly provide your train and test dataset?

Hello, I was wondering what were your hw specs for training/inference.

• How much memory or what GPU is required for training/testing?
• How much time does it take to make inference?

The paper says "we have faster inference time", do you have some numbers?

Thanks.

From the paper 3.4 Loss function, the total loss function contains norml loss and boundaries loss. Considering you use the pre-training model to generate normal and boundaries image before training, the loss function should not contians thest two parts. How can I understand that?

Hi everyone,

In the recent one month, I have received a lot of emails and messages regarding to the missing datasets issue. After digging in my lab servers, PC, old laptops and so on, I eventually found a copy of "pre-processed (resize operation)" ground truth data of Matterport3D as well as ScanNetv2. After that, I contacted Yinda Zhang for the news but did not receive any reply for a while. As it can be used for academic advancement, I decided to share them although it might be a bit different from the original data.

In the following 7-14 days, I will organize these data, upload to a "safer place", and publish rules for using datasets on our GitHub. Also, feel free to contact me if you own an "original copy". I will close this issue once I finish it.

Thanks for your patience.

There is a bad news, the data set of yindaz/DeepCompletionRelease has been lost, I also sent them an email to apply for the data set, but did not reply for a long time. So could you please send me the dataset of DeepCompletionRelease, or the render depth, estimated normal and estimated boundary of DeepCompletionRelease that you used in your project. Otherwise I cannot reproduce your work.
https://github.com/yindaz/DeepCompletionRelease/issues/45#issue-993195123

Hi,

I see in this work a boundary consistency network is proposed to predict boundary, but I wonder why not extracting boundary directly by the Sobel filter. This is more intuitive and much simpler, right? Are there some explanations?

thank you for your instructive work.

Hi
It is difficulty for me to follow yindaZ's repository to get render depth, estimate normal and estimate boundary.
Could you kindly provide your prepared dataset in Training / Testing?

could you share the training dataset?