MendNet: Restoration of Fractured Shapes Using Learned Occupancy Functions

Simple item page
dc.contributor.authorLamb, Nikolasen_US
dc.contributor.authorBanerjee, Seanen_US
dc.contributor.authorBanerjee, Natasha K.en_US
dc.contributor.editorCampen, Marcelen_US
dc.contributor.editorSpagnuolo, Michelaen_US
dc.date.accessioned2022-06-27T16:19:52Z
dc.date.available2022-06-27T16:19:52Z
dc.date.issued2022
dc.description.abstractWe provide a novel approach to perform fully automated generation of restorations for fractured shapes using learned implicit shape representations in the form of occupancy functions. Our approach lays the groundwork to perform automated object repair via additive manufacturing. Existing approaches for restoration of fractured shapes either require prior knowledge of object structure such as symmetries between the restoration and the fractured object, or predict restorations as voxel outputs that are impractical for repair at current resolutions. By leveraging learned occupancy functions for restoration prediction, our approach overcomes the curse of dimensionality with voxel approaches, while providing plausible restorations. Given a fractured shape, we fit a function to occupancy samples from the shape to infer a latent code. We apply a learned transformation to the fractured shape code to predict a corresponding code for restoration generation. To ensure physical validity and well-constrained shape estimation, we contribute a loss that models feasible occupancy values for fractured shapes, restorations, and complete shapes obtained by joining fractured and restoration shapes. Our work overcomes deficiencies of shape completion approaches adapted for repair, and enables consumer-driven object repair and cultural heritage object restoration. We share our code and a synthetic dataset of fractured meshes from 8 ShapeNet classes at: https://github.com/Terascale-All-sensing-Research-Studio/MendNet.en_US
dc.description.number5
dc.description.sectionheadersLearning and Creating
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume41
dc.identifier.doi10.1111/cgf.14603
dc.identifier.issn1467-8659
dc.identifier.pages65-78
dc.identifier.pages14 pages
dc.identifier.urihttps://doi.org/10.1111/cgf.14603
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf14603
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectCCS Concepts: Computing methodologies --> Shape analysis; Neural networks
dc.subjectComputing methodologies
dc.subjectShape analysis
dc.subjectNeural networks
dc.titleMendNet: Restoration of Fractured Shapes Using Learned Occupancy Functionsen_US
Files
Original bundle
Now showing 1 - 3 of 3
Thumbnail Image
Name:
v41i5pp065-078.pdf
Size:
16.35 MB
Format:
Adobe Portable Document Format
Download
Thumbnail Image
Name:
sgp_2022_camera_ready_supplementary.pdf
Size:
9.38 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
sgp_2022_camera_ready_video.mp4
Size:
71.44 MB
Format:
Unknown data format
Download
Collections
41-Issue 5