The goal of this project is to develop a mixed reality sketching application for architectural design. The application should integrate novel semantic sketching techniques with machine learning and reasoning algorithms. Machine Learning will be used to recognize sketched objects and also to retrieve design inspirations (sketches, images and 3d models). Furthermore, personalized style transfer of sketches to library objects should provide a consistent aesthetic visualization of architectural sketches. Integration of semantic information via ontologies and logic reasoning will derive machine understanding of sketched design intentions. Sketched objects need to be converted into semantically enriched parametric models, which will be used in other applications of the SFB. Feedback from simulation results need to be integrated into the sketching application to aid the evaluation of design goals.
Necessary: Master in computer science (machine learning and/or visual computing); C# programming skills inside the Unity environment.
Desired: Experience with Reasoning and Machine Learning, preferable inside Unity.
Necessary: Capacity for teamwork in a large multi-, inter- and transdisciplinary research team (architecture, building engineering, mathematics, computer sciences); Ability to explain complex contents to a non-specialized public in simple words; Excellent English skills.
Desired: Basic knowledge of creative means of expression; Comprehensive education.
The department of “Integrated Planning and Industrial Building” at the faculty of Civil Engineering offers interdisciplinary teaching and research platform, providing teaching for students of both Civil Engineering and Architecture, thus offering the opportunity to evaluate the developed tools in interdisciplinary design courses. Further on, a number of BIM tools and licenses is available, as well as the strong connection with industry and practice. We offer a friendly, supporting working environment.