Recognition of the emotion in abstract expressionistic paintings using a machine learning approach

The topic of the research described in this master thesis is: “A machine learning approach to emotion classification of abstract expressionistic paintings.” Emotion in paintings is a result of the sum off all the elements within a painting. In abstract expressionist paintings abstract shapes, expressive colors and textures depict emotion in the painting. The research involves the abstract expressionistic art movement, evaluation of human emotions and machine learning algorithms. The main points of interest within this research is the relation between abstract expressionistic paintings and the emotion they are assumed to evoke at human observers. This leads to the following research question: “Is it possible to predict how people would react emotionally when observing an abstract expressionistic painting?”