Computational Creativity Under the Framework of Recoloring Art Paintings for Color-Blindness

How creative is the machine-learning based modification of digitized art paintings in enhancing the color perception for audiences with color vision deficiencies (CVD) by altering color contrast and forms? The current study investigates the above research question by integrating computational creativity theory, aesthetic theory, and machine learning to examine the sensory perception of the modified artworks. The modification process is seen as a creative condition that transforms the framework, space, and perception of the original artwork. The paper also explores existing creativity theories that focus on the stages of the creative process within specific activities, describing the regime of artistic functioning that serves as a type of identification inherent to art, reshaping the relationships between its forms. The resulting theoretical analysis deduces a questionnaire that summarizes the basic facets of creativity involved in the current inquiry. To conduct the imminent testing and evaluation procedure, three algorithms are employed to perform the recoloring of six art paintings with respect to two types of CVDs namely, protanopia and deuteranopia. Several people related to culture, the arts, psychology, and computer science participated in evaluating the creative behaviors of the three algorithms. Finally, statistical analysis was carried out to identify certain directions of the above behaviors.