no code implementations • 29 May 2023 • Xinchao Song, Nikolas Lamb, Sean Banerjee, Natasha Kholgade Banerjee
We also introduce Fantastic Breaks Imaged, the first large-scale dataset of 11, 653 real-world images of fractured objects for training and evaluating image-based shape repair approaches.
no code implementations • CVPR 2023 • Nikolas Lamb, Cameron Palmer, Benjamin Molloy, Sean Banerjee, Natasha Kholgade Banerjee
Automated shape repair approaches currently lack access to datasets that describe real-world damaged geometry.
1 code implementation • 22 Nov 2022 • Nikolas Lamb, Sean Banerjee, Natasha Kholgade Banerjee
We generate a high-resolution restoration shape by inferring a corresponding complete shape and a break surface from an input fractured shape.
1 code implementation • 11 Oct 2022 • Nikolas Lamb, Sean Banerjee, Natasha Kholgade Banerjee
We represent the occupancy of a fractured shape as the conjunction of the occupancy of an underlying complete shape and the fracture surface, which we model as functions of latent codes using neural networks.