Shaping the aesthetical landscape by using Image statistics measures
Description
• The data collection of the 11-dimensional aesthetic space, constructed with the sample of artistic triangles, has never been realized in the experimental aesthetics literature. In this research, we offer data that contribute to the understanding of the aesthetic organization of such a space. • This aesthetic investigation through the analysis of the 11 developed metrics sheds light on the internal organization of the aesthetic space. The results of this investigation contribute to the evaluation of broad regions of such space where elements of low, medium, and high aesthetic value reside • The data provided can be used to implement automated aesthetic evaluation systems in both artistic and manufacturing contexts. We provide five data sets: (a) triangles made by artists; (b) the distribution of triangles as a function of 11 aesthetic metrics (55 images), provided as a 3D application visualizing data as a catalogue; c) the plots of the calculated Smooth Kernel Distribution, on two dimensions (55 images); d) the 3d plots of the Smooth Kernel Distribution calculated in 3 dimensions (110 images), provided as navigable environments in the software Mathematica; e) aesthetical values for the 83 triangles for all the 11 metrics developed.
Files
Steps to reproduce
83 artistic objects were developed by students attending the 3rd year of the Fine Arts course (Artistic_Triangles_Dataset). We asked ourselves. Can these objects be evaluated automatically, in order to extract useful values for their aesthetic classification? Starting from this problem, we developed 11 different metrics, which allowed us to create an 11-dimensional aesthetic space, within which we investigated how triangles are distributed, using intersections of two metrics at a time (2D_Triangle_Distribution_Dataset). the metrics are as follows: Birkhoff's Measure, Euclidean Distance, Entropy, Fractal dimension, Morphological Component, MorphologicalEulerNumber, Image Key points, Corners, Image Lines, 0 dim Mesh number, 1 dim Mesh number. The values of these metrics for each artistic triangle are provided in the file Triangles_Aesthetic_Values.csv. Then using the Probability Density Function (PDF), we crossed 2 metrics at a time with the PDF, producing 3D plots. Images of these plots are provided in the Probability_Distribution_Function_Data file. Visualisations of such probability spaces were also made, crossing 3 metrics at a time and the PDF. A sample of such data can be seen in the Mathematica Sample of 3D_Data.nb file. To account for the aesthetic properties of the triangles, we then created a navigable museum environment in which we displayed all the triangles, for more accurate aesthetic enjoyment. The link to this application can be found at the following web address: https://armonicamente.unical.it/nuova-sezione-mostra/. Another interactive application allows users to look at the triangles one by one, as if browsing through a catalogue. This application can be found at the following website: https://armonicamente.unical.it/catalogo-triangoli/. The reference article for this work is the following: Bertacchini F., Pantano S:P., Bilotta E. Shaping the aesthetical landscape by using Image statistics measures. Acta Psychologica, In Press.