fabrication-to-ai-renderings

Sara Codarin

The research explores the interplay between robotic fabrication and AI-assisted image rendering. AI technologies have simplified the process of generating images through text-to-image AI, image-to-image blends, and hybrid operational modes. However, limitations exist in the constraints imposed by the two-dimensional image boundaries, which may provide an initial atmosphere or early-stage design idea that does not fully incorporate three-dimensional intelligence. Additionally, the inconsistency in image variations hampers the ability to refine and iterate on a specific concept. Addressing these issues, the study aims to integrate insights from fabrication-aware model creation into the AI image generation process.


LTU Research Day

Stable Diffusion workflow

Stable Diffusion ComfyUI renderings

Foam model

Stable Diffusion ComfyUI rendering

Robotic fabrication-to-AI renderings

Depth masks

Stable Diffusion ComfyUI renderings

Robotic hot wire cutting

Robotic hot wire cutting