hongxiaoy / ISO

[ECCV 2024] Monocular Occupancy Prediction for Scalable Indoor Scenes
https://hongxiaoy.github.io/ISO
Apache License 2.0
29 stars 1 forks source link

2 processing files to get the occ-scannet #3

Closed YkiWu closed 2 months ago

YkiWu commented 2 months ago

Thank you for introducing the Occ-SanNet dataset in your paper Monocular Occupancy Prediction for Scalable Indoor Scenes. It has provided valuable new data for my research. While following the data acquisition and processing steps provided in the link https://huggingface.co/datasets/hongxiaoy/OccScanNet, I encountered an issue. I was unable to locate the Python files preprocess_gt.py used in Step 6 (Reformulate CompleteScanNet) and generate_gt.py used in Step 7 (create the OccScanNet). Could you please advise me on where I can obtain these two processing files?

hongxiaoy commented 2 months ago

Thanks. I updated the preprocess_gt.py and generate_gt.py to the provided huggingface link, you can have a look at it.

And the README on huggingface repo is also updated, you can follow it to generate dataset by yourself.

For simplifications, we have provided the only data you may use, you can switch to use it instead. But it's good to try generating by yourself again to make sure the process is all right.

Thanks for you question! If you still faced with some problems, just feel free to ask me!

YkiWu commented 2 months ago

Thanks. I updated the _preprocessgt.py and _generategt.py to the provided huggingface link, you can have a look at it.

And the README on huggingface repo is also updated, you can follow it to generate dataset by yourself.

For simplifications, we have provided the only data you may use, you can switch to use it instead. But it's good to try generating by yourself again to make sure the process is all right.

Thanks for you question! If you still faced with some problems, just feel free to ask me!

Thanks very much!!!