Grants and Contributions:

Grant or Award spanning more than one fiscal year. (2017-2018 to 2018-2019)

The primary objective of this project is prediction of a continuous output Y from images acting as covariates Xx000D
through a nonlinear regression function f() in the equation: Y = f(X) + residual. In order for the regressionx000D
model to be valid all of the X covariates must have the same dimension. Unlike in most image processingx000D
applications images in this project are of differing sizes. In standard applications irregularly sized rectangularx000D
images can be transformed via compression, stretching, cropping, and padding to bring the images to ax000D
common size. However in our application a transformed image with a high value of Y will be equivalent to anx000D
un-transformed image of the same dimension with a low value of Y. In other words, the size normalizingx000D
transformation process induces artifacts that will destroy the regression relationship. This project expands thex000D
methods used for image processing to account for this type of problem. Ultimately the insights of thex000D
regression model will be used to predict optimal ways in which image X can be modified so as to produce anx000D
increased/decreased value of Y. In this project we try to predict year Y images X from an art database wherex000D
art pieces are arbitrarily sized rectangles with any aspect ratio. Compressing an image will distort the imagex000D
contents in ways that could be inconsistent with its time period. Cropping an image removes importantx000D
features for prediction. Padding images with extrapolated coloured space makes the image equivalent to ax000D
different art piece. In such cases an early renaissance piece may become dominated by the transformationx000D
features making it equivalent to a recent modern art piece. The secondary goal of this work is to examine howx000D
to modify an art work so as to make it equivalent to an earlier or later form of that style. These methods arex000D
applicable to the partner company, Unbounce, in that they will assist them in advising their client customers inx000D
how to modify artwork optimally while maintaining it's overall style.