Bitte aktualisieren Sie Ihren Browser zur korrekten Anzeige dieser Webseite.

Design Research

Elements of APN – Using Machine Learning for the Detection, Typologization, and Comparison of Iconographic Elements in Visual Artifacts.

Research set-up project

Andreas Koller, Lecturer, Visual Communication, Global Design

The project combines machine learning and manual research to undertake a formal, visual analysis of the «APN» print work cycle (1953/54) by the Jikken Kōbō collective, as well as Japanese art and design magazines from the late 1940s and early 1950s.

The project combines machine learning and manual research to undertake a formal, visual analysis of the APN print work cycle (1953/54) by the Jikken Kōbō collective, as well as Japanese art and design magazines from the late 1940s and early 1950s. Following this visual examination, the study will be situated within a broader design and art historical context. Both the analysis – conducted in collaboration with the IVIA Lab at ETH Zurich – and its contextualization are part of a case study within the framework of a publication project, funded by an IfCAR Grant from ZHdK, and a larger research initiative aimed at developing a subsequent funding proposal for third-party grants.

APN works extracted from Asahi Graphu (left), reference dataset: digital facsimiles of Japanese art and design magazines from the late 1940s and early 1950s (right).
APN works extracted from Asahi Graphu (left), reference dataset: digital facsimiles of Japanese art and design magazines from the late 1940s and early 1950s (right).
Neighborhoods of Similar Visual Artifacts
Neighborhoods of Similar Visual Artifacts
Dashboard: See What the Model Sees (Heat Maps)
Dashboard: See What the Model Sees (Heat Maps)