Geometry-Consistent Neural Shape Representation with Implicit Displacement FieldsDownload PDF

Sep 29, 2021 (edited Feb 10, 2022)ICLR 2022 PosterReaders: Everyone
  • Keywords: implicit functions, shape reconstruction, shape representation, object reconstruction
  • Abstract: We present implicit displacement fields, a novel representation for detailed 3D geometry. Inspired by a classic surface deformation technique, displacement mapping, our method represents a complex surface as a smooth base surface plus a displacement along the base's normal directions, resulting in a frequency-based shape decomposition, where the high-frequency signal is constrained geometrically by the low-frequency signal. Importantly, this disentanglement is unsupervised thanks to a tailored architectural design that has an innate frequency hierarchy by construction. We explore implicit displacement field surface reconstruction and detail transfer and demonstrate superior representational power, training stability, and generalizability.
  • One-sentence Summary: We extend classic displacement mapping to the neural implicit framework. The resulting novel implicit representation demonstrates superior reconstruction accuracy, parameter efficiency and enable implicit shape editing such as detail transfer.
  • Supplementary Material: zip
10 Replies