*19xx in xxx, lives and works in Munich (DE)
*19xx in xxx, lives and works in xxx
In cooperation with Google Arts & Culture
X Degrees of Separation
2017, Multi media Installation
X Degrees of Separation uses machine learning to trace the relationships between 750,000 exhibits from the around the world, enabling the viewer to discover new connections between works. The underlying question is how we define the relations between artworks and artefacts.
Art history frequently orders works in a linear progression of time or by country of origin. X Degrees of Separation focuses on the visual components of artefacts using machine learning, that is, by encoding visual representations into numerical data and drawing pathways between them.
Using convolutional neural networks of 128 dimensions, the image features are captured based on similarity of color, texture, shape and other complex features that initially cannot be labelled. These features are compared for their similarities and grouped accordingly on a graph. By comparing the distances of each feature vector on the graph, the relationship between all artifacts in the dataset can be measured.
Bridging art historical periods, movements, styles and techniques, X Degrees of Separation tries to connect formal and conceptual features of artworks in an attempt to draw new connections across history and art. The piece raises questions how of knowledge is drawn from artworks and their contextual framing.