Fetching the paper…

Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentation · Around