Topic outline

  • A few thematic project topics. Students are free to edit these and propose their own topics. Re-scoping needed. The list will be updated as the course progresses based on seminar talks.

    • A user interface for routing design for optimizing stroke ambulance siting, with the goal of minimizing response times.
    • A workflow for prompt-based generative AI systems (large language models) for interactive scenario/storyboard building. A case example to be included (w/ efficacy assessment?)
    • A workflow proposal for physical product design, involving (prompt based) generative AI tools.
    • An empirical evaluation of emerging LLM-based (large language model) tools for Photoshop  
    • A literature study of generative AI methods in field of X, where X is your design field. 
    • A workflow or interface proposal for human-in-the-loop optimization methods in X, where X is your design field. (HITL methods: see in particular Bayesian optimization)
    • A designer study of a Figma plug-in with AI capabilities (ask Antti)
    • A review of computational methods for assisting graphical UI design 
    • A user interface for using GPT-3 / Stable Diffusion for a creative task (you define).
    • A study of designers' starting to use AI (e.g., GPT-3, DALL-E, Stable Diffusion)
    • An evaluation of 3D assets (3d models) generated by AI (e.g., the Nvidia example Perttu showed)
    • Using LLMs (large language models like GPT-3, Dalle2) for evaluating user experience (without real users)
    • Social simulacra and their use in design (what, how?) 
    • Simulating individual differences among users with AI (where, how?)
    • Simulation studies of interior design using Virtual Home (http://virtual-home.org)
    • VR/AR experiencing of 3D models
    • Collaborations support for VR/AR in computational design
    • Helsinki city model (search for "Forum Virium") and its use for the design of the built environment
    • LLMs for ideating early-stage architectural designs
    • Scale models and 3D printing: support from computational design methods