Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view Human Reconstruction.
Tong HeJohn P. CollomosseHailin JinStefano SoattoPublished in: NeurIPS (2020)
Keyphrases
- single view
- multi view
- single view reconstruction
- perspective projection
- three dimensional
- multiple views
- co training
- surface reconstruction
- vanishing points
- view invariant
- high resolution
- visual hull
- machine learning
- repetitive structures
- polyhedral objects
- user input
- reconstruction method
- x ray images
- image reconstruction
- learning process
- smooth surfaces
- back projection
- calibrated cameras
- depth map