A Dynamic Approach for the Generation of Perceptual Associations

Authors

Abstract

We propose a dynamic approach and computational artefact for the generation of cross-modal associations between the domains of visual communication and perception of emotions. This is accomplished by applying three key steps: (i) assemble an alphabet (ii), define a shape grammar, and (iii) generate computationally a dynamic representation for the previous point. We restrain the visual expression to a line movement and limit our set of emotions to four in order to establish an initial solid baseline of experiments. Experiments were done to assess the validity of the established associations be- tween emotions and visual features. Results have con- firmed some associations previously suggested by literature and have emphasized others. This approach may be a key aspect to solve problems of multi-domain con- texts where non-verbal communication is required to connect two or more domains, e.g. visual arts, artistic performances, non-verbal languages, and entertainment.

Conference

International Conference on Computational Creativity (ICCC) 2019