Pre-trained Models (Neurips24)

This release contains the pre-trained QUEEN models generated from the N3DV dataset. These are the models that produced the numbers in our NeurIPS 2024 paper.

The models are split into multiple files for upload. Please follow these steps to download and extract the models:

1. Download all parts of the model files.
2. Concatenate the files and extract them using the following commands:

cat queen_neurips_cook_spinach.* > queen_neurips_cook_spinach.tar.gz
tar -xzvf queen_compressed_cook_spinach.tar.gz

Each extracted folder contains a frames directory with .ply point cloud files for each frame. You can render these frames using either render.py or render_fvv.py.
Model license terms: Creative Commons — Attribution-NonCommercial-ShareAlike 4.0 International — CC BY-NC-SA 4.0.

Pre-trained Models (Compressed)

This release contains the pre-trained QUEEN models generated from the N3DV dataset. These models were generated with a corrected version of the code, and contains both dense .ply files and compressed .pkl files.

The models are split into multiple files for upload. Please follow these steps to download and extract the models:

1. Download all parts of the model files.
2. Concatenate the files and extract them using the following commands:

cat queen_compressed_cook_spinach.* > queen_compressed_cook_spinach.tar.gz
tar -xzvf queen_compressed_cook_spinach.tar.gz

Each extracted folder contains a frames directory with .ply point cloud files for each frame. You can render these frames using either render.py or render_fvv.py.
Model license terms: Creative Commons — Attribution-NonCommercial-ShareAlike 4.0 International — CC BY-NC-SA 4.0.