Kind Request about the Training Data.

[OliverHxh]

Thank you for your outstanding work.

From the description in zd11024/Video-3D-LLM_data, I noticed that you provide most of the necessary data for training. However, we still need to download the scans data from ScanNetV2, which is a massive dataset that requires several days to download. Additionally, I found that the scans data is not directly usable for training and requires further processing. Most of this processed data has been provided on HuggingFace, except for the posed_images data.

Therefore, I was wondering if you could kindly provide the download link for the posed_images data. This would significantly reduce the time needed to run your code and further explore your remarkable work.

Thank you once again for your time and effort.

[zd11024]

I have uploaded the processed images to the huggingface link.

[OliverHxh]

Thanks for your help!

[zd11024]

You can concatenate these files through cat posed_images_part* > posed_images.tar.gz.

[OliverHxh]

Thank you. You are very helpful.