Closed Gofinge closed 2 years ago
Thanks for the note and recommendation. Unfortunately such data is not currently available with the first release of ARKitScenes partly due to storage consideration as those assets are pretty large. However, we are considering the release of the raw Faro point clouds in future and will keep you posted.
Closing the issue for now. Feel free to re-open it if you have further questions.
Re-opening the issue till fully resolved.
When can you release the raw Faro point clouds?
@afshindn @PeterZheFu thanks for the comprehensive dataset! I'm interested in the Faro point clouds as well, would it be possible to provide a few samples and any other information about the scanning process, such scanner settings, resolution and scan time? This would be useful for us to evaluate on custom datasets.
Hi, thanks again for the interest in our dataset!
We recently released high resolution venue point-clouds that were produced by Faro Lidars. They are now available to download from the raw dataset by using the --download_laser_scanner_point_cloud
flag.
Please see the data documentation and raw dataset documentation for more information.
Hi, really thanks for valuing our request. I was trying to download a high-resolution venue by following the example script in the image shown above. And It seems that "Laser scanner point clouds for video 41048190 are not available" and the video id is provided by docs. @arik1089
Hi, yea you are right @Gofinge, thanks for helping with a quick fix to the docs.
I have downloaded the video_id 47333462 (laser scanner point cloud), but I found that the name of the downloaded point cloud file is different from 47333462,and its name is 467138.
Hi, @zbs881314 : The laser_scanner_point_clouds_mapping.csv in the ARKitScenes dataset home folder might solve your question. A mapping exists between video_id and visit_id.
Thank you for the impressive large-scale indoor dataset collecting work. It's a significant dataset with many possibilities for high-level application scenarios, and I like it.
It is thoughtful to generate ground truth high-resolution depth maps by discarding geometry which a direct line-of-sight from the novel viewpoint cannot be guaranteed.
But I think if we can get access to raw Faro high-resolution XYZRGB point cloud, it will grant the dataset more potential, such as point cloud completion task.
Will it be possible for us to get access to much more raw data collected in your well-designed data collecting process? We can explore more meaningful settings to explore 3D understanding in such a large-scale real indoor dataset.