Mean Face (μ)
Top-20 Eigenfaces
—
Explained Variance
■ per component
— cumulative
How Variance is Computed
1. Center data
2. SVD
3. Eigenvalue
4. Total var
5. Ratio
Sample Faces (click to select)
Input
—
=
Mean Face (μ)
+
Reconstruction
—
+
Residual
—
20
—
Progressive Reconstruction
Eigenface Coefficients click row to inspect