Start Date
9-5-2019 10:30 AM
End Date
9-5-2019 11:45 AM
Document Type
Event
Keywords
computer vision, deep style, architecture
Description
In this paper, we hope to demonstrate approaches intended to further direct an implementation of Deep Style AI to create architectural works. We explore several processes through an interdisciplinary lens that are used in tandem to direct the AI to complete a controlled task using a human in the loop scenario. Through an understanding of the mechanics of the Deep Style AI in action, we can interrupt both the AI’s learning and application processes. By interrupting these processes, we can insert intentional data used to direct the work of the AI. By demonstrating the AI’s ability to complete a controlled and directed task we hope to outline the possibility that this readily available and open source neural network can assist in the formation of original creative works.
DOI
https://doi.org/10.5038/EGPI4882
A Multidisciplinary Approach to AI Generated Architecture
In this paper, we hope to demonstrate approaches intended to further direct an implementation of Deep Style AI to create architectural works. We explore several processes through an interdisciplinary lens that are used in tandem to direct the AI to complete a controlled task using a human in the loop scenario. Through an understanding of the mechanics of the Deep Style AI in action, we can interrupt both the AI’s learning and application processes. By interrupting these processes, we can insert intentional data used to direct the work of the AI. By demonstrating the AI’s ability to complete a controlled and directed task we hope to outline the possibility that this readily available and open source neural network can assist in the formation of original creative works.