objaverse-xl
objaverse-xl copied to clipboard
allenai
🪐 Objaverse-XL is a Universe of 10M+ 3D Objects. Contains API Scripts for Downloading and Processing!
## Alignment Fine-tuning Annotations Is this just a subset of objaverse-xl that's ensured to be aligned with text description?
Thanks for the amazing work! But i have a problem when downloading the dataset: ``` import objaverse.xl as oxl annotations = oxl.get_annotations( download_dir="~/.objaverse" # default download directory ) annotations print(annotations)...
This PR integrates Cap3D captions into Objaverse annotations along with point-clouds download link. `objaverse.load_uids()` will first download a [`cap3d_captions.json.gz`](https://huggingface.co/datasets/tiange/Cap3D/blob/main/Objaverse_files/cap3d_captions.json.gz) file locally and then assign captions to each inquired `UID`. For...
thank you for your work, I tried to upload data to s3, but I had an error hanging all the time, what could this be related to?
I was able to get images of all the objects from Objaverse 1.0. Is there any source for thumbnails of all the objaverse-XL objects?
Hello! Can I use your dataset in another research and cite your paper? What is your policy of dataset using?
If so, how could one filter those objects that are in Objaverse-XL but not in 1.0?
Hi, In the objaverse-xl paper, it refers to this the 3D assets in this dataset coming with metadata such as textual pairs, aesthetic scores, NSFW predictions, etc. Will this metadata...
Do you have some numbers on the Objaverse-XL size once downloaded ? (with 3D objects) ? What impacts does github repo have ? with zip/tar/tar.gz/files ?
The scaling graph for PixelNeRF look very interesting. Will you plan to release the checkpoint and implementation for the PixelNeRF model?
