search

alexklwong / calibrated-backprojection-network

PyTorch Implementation of Unsupervised Depth Completion with Calibrated Backprojection Layers (ORAL, ICCV 2021)
Other
117 stars 24 forks source link

about NYUv2 data #9

Closed lqzhao closed 2 years ago

lqzhao commented 2 years ago

Hi Alex, I notice that you update the script for downloading the NYUv2 dataset. Thanks. I downloaded the raw data from the NYUv2 official website weeks ago. But I found the unzipped data contains many files, with extensions like: '.dump', '.pgm', and '.ppm'. And the name of files are like: image

However, the setup python file of NYUv2 seems to only accept the files with '.png' extension. https://github.com/alexklwong/calibrated-backprojection-network/blob/73e3943169b3baf0e5b60e1b3378337245a03464/setup/setup_dataset_nyu_v2.py#L246

So my question is: Did I download the correct NYUv2 data? or How can I set up the data for your code? Thanks in advance!

alexklwong commented 2 years ago

Ah I just realized that our lab's internal copy of NYUv2 has been post processed and synced. I will write a setup to do that post processing and synchronization. That might be valuable for anyone who wants to work on NYUv2.

lqzhao commented 2 years ago

Ah I just realized that our lab's internal copy of NYUv2 has been post processed and synced. I will write a setup to do that post processing and synchronization. That might be valuable for anyone who wants to work on NYUv2.

That's great, thank you for your effort.

alexklwong commented 2 years ago

Sorry this took so long, we will upload the preprocessed version of NYUv2 onto Google drive.

alexklwong commented 2 years ago

@lqzhao here are the new instructions for setting up NYUv2. We've synchronized and aligned the depth and image frames and stored it into a zip. You can follow the README to download and set up the dataset.

https://github.com/alexklwong/calibrated-backprojection-network/tree/nyu-v2-setup

This will be merged into master shortly.

lqzhao commented 2 years ago

Thank you~