DeepSDF: Learning Continuous Signed Distance Functions for ...
DeepSDF: Learning Continuous Signed Distance Functionsfor Shape RepresentationJeong Joon Park1,3Peter Florence2,3Julian Straub3Richard Newcombe3Steven Lovegrove31University of Washington2Massachusetts Institute of Technology3Facebook Reality LabsFigure 1:DeepSDF represents Signed Distance Functions (SDFs) of shapes via latent code-conditioned feed-forward decoder images are raycast renderings of DeepSDF interpolating between two shapes in the learned shape latent space. Best viewed graphics, 3D computer vision and roboticscommunities have produced multiple approaches to rep-resenting 3D geometry for rendering and provide trade-offs across fidelity, efficiency and com-pression capabilities. In this work, we introduce DeepSDF,a learned Continuous Signed Distance Function (SDF) rep-resentation of a class of shapes that enables high qual-ity shape representation, interpolation and completion frompartial and noisy 3D input data. DeepSDF, like its clas-sical counterpart, represents a shape s surface by a con-tinuous volumetric field: the magnitude of a point in thefield represents the Distance to the surface boundary and thesign indicates whether the region is inside (-) or outside (+)of the shape, hence our representation implicitly encodes ashape s boundary as the zero-level-set of the learned func-tion while explicitly representing the classification of spaceas being part of the shapes interior or not.
Note that the binary occu-pancy function is a special case of SDF, considering only ‘sign’ of SDF values. As DeepSDF models metric signed distance to the surface, it can be used to raycast against sur-faces and compute surface normals with its …
Download DeepSDF: Learning Continuous Signed Distance Functions for ...
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document: