Using Subspace Multiple Linear Regression for 3D Face Shape Prediction from a Single Image.
Mario CastelánGustavo A. Puerto SouzaJohan Van HorebeekPublished in: ISVC (2) (2009)
Keyphrases
- single image
- multiple linear regression
- shape recovery
- partially occluded
- d morphable model
- regression model
- software effort
- multiple images
- input variables
- software reliability
- d scene
- intrinsic images
- principal component analysis
- spherical harmonics
- light source
- shape from shading
- super resolution
- shape model
- three dimensional
- information from multiple images
- low dimensional
- facial expressions
- long term
- high dimensional
- feature space
- feature points
- dimensionality reduction
- high resolution
- feature extraction
- range images
- surface normals
- linear subspace
- artificial neural networks
- multiscale
- computer vision