Coefficients for Face Recognition Experiments
To allow an objective comparison of different face recognition algorithms, we provide model coefficients for several different public available image databases. The coefficients are obtained with our state of the art fitting algorithms. We hope that other researchers developing future algorithms based on the BFM will likewise provide the coefficients to enable more precise algorithm comparison and accuracy analysis.
Public Face Image Databases
- feretR2b_fres.mat contains the coefficients for FERET images.
- pie_lights_fres.mat contains the coefficients for CMU - PIE images.
Public 3D Face Database
- und_fres.mat contains the fitting results for the UND database. (coming soon)