Figure Prediction

Bilge Arslan

2023 Spring
BS723 | Machine Learning Applications in Architecture


Depending on the nature of the project or the setting, it can take a long time for the artist to take a look at the pictures that the human tree or other comparable objects, which are frequently used in architectural drawings and visualizations, demand. We have been seeking for images with particular forms and sizes to particular locations while applying images in visualizations.
The goal of this machine learning model was to create a quicker solution than the time-consuming manual search through human figures. It builds a library of a predetermined number of human
figures and learns with variously warped versions of each figure. Thus, the machine will be able to provide the user with the human figure that most closely resembles the abstract figure they have drawn when they draw a human figure anyway they choose.

Figure Predictionadmin