lecture7comp

Separating Style and Content with Bilinear ModelsJoshua B. Tenenbaum, William T. FreemanComputer Examples: 

Separating Style and Content with Bilinear Models Joshua B. Tenenbaum, William T. Freeman Computer Examples Barun Singh 25 Feb, 2002

Slide2: 

PHILOSOPHY & REPRESENTATION Data contains two components: style and content Want to represent them separately

Slide3: 

PROBLEMS TO BE SOLVED Given a labeled training set of observations in multiple styles and content classes, Fit asymmetric model (find A and b for known styles and contents) using SVD Extrapolate using the estimated style matrix OLC used to solve overfitting problem Parameters involved: l l = 0 : Purely asymmetric model l =  : Purely symmetric model extrapolate a new style to unobserved content classes

Slide4: 

PROBLEMS TO BE SOLVED Given a labeled training set of observations in multiple styles and content classes, classify content observed in a new style Fit asymmetric model Select content class c that maximizes Pr(s’,c|y)

Slide5: 

PROBLEMS TO BE SOLVED Given a labeled training set of observations in multiple styles and content classes, translate from new content observed only in new styles into known styles or content classes Fit symmetric model (find W, a, and b for known styles and contents) using iterated SVD procedure

Slide6: 

TOY EXAMPLE - intro Image made of 4 pixels, each of which are either white or red. Style represents if the top or bottom rows are red or white Content represents if the left or right columns are red or white. SYMMETRIC MODEL

Slide7: 

TOY EXAMPLE - intro ASYMMETRIC MODEL *Note: Images drawn as blocks, but represented as vectors, not matrices

Slide8: 

TOY EXAMPLE - extrapolation

Slide9: 

FONTS EXAMPLE - extrapolation

Slide10: 

FONTS EXAMPLE - extrapolation

Slide11: 

TOY EXAMPLE - classification

Slide12: 

TOY EXAMPLE - classification 2: Use Separable Mixture Model w/ EM to classify

Slide13: 

FACES EXAMPLE - translation

finito: 

finito