Victor Sardenberg

In this seminar, the current possibilities of artificial intelligence models for the generation, manipulation and analysis of architectural drawings are examined. Over the past half a decade, the architectural images generated by AI have developed from the edges of experimental architecture into a research topic that is popular around the world. This seminar interfaces with several AI types that democratizes machine learning for artists and designers. Several topics and techniques are explored: style transfer, generative adversarial networks for image generation, photogrammetry, image-to-text and text-to-image models, object recognition and image segmentation.