The course will address various topics within the field digital imagery and CG production. By introducing advanced tools and techniques and through productive experimentation we will also address the ongoing discussion around the digital realm, image sciences and the iconic turn. Students are invoked to further develop their technical abilities as well as define/refine/redefine an attitude towards their style of visual communication. We will also see the strong relation of DCC-production to our physical space, tools and effects.

CGI production overlaps the domain of architectural production although approaches and methods are slightly different. Where architects need to measure and keep a consistency between the existing physical space and the artifact, CGI production just operates in the realm of visual content.  CGI tools are more derived from sculptors painters photographers animators than from constructors or engineers.  CGI production needs to be fast, imaginative- consistency is needed just when you can see it. Think of a set design from the movies.

SHEDULE

01: 07 October 2025, 10:45–12:15 Seminar Room 10, First Meeting
02: 14 October 2025, 10:45–12:15 Seminar Room 10
03: 21 October 2025, 10:45–12:15 Seminar Room 10
04: 28 October 2025, 10:45–12:15 Seminar Room 10
05: 04 November 2025, 10:45–12:15 Seminar Room 10, Mini Research Task presentation & Deadline Basic Tutorial
06: 11 November 2025, 10:45–12:15 Seminar Room 10, Mini Research Task presentation
07: 18 November 2025, 10:45–12:15 Seminar Room 10
08: 25 November 2025, 10:45–12:15 Seminar Room 10
09: 02 December 2025, 10:45–12:15 Seminar Room 10
10: 09 December 2025, 10:45–12:15 Seminar Room 10
11: 16 December 2025, 10:45–12:15 Seminar Room 10

Interesting Stuff

About AI image to model

Mario Carpo lecture at Harvard GSD

00 Mini Research Task: “AI in Computer Graphics for Architecture”

Context

Artificial Intelligence is rapidly reshaping how we imagine, model, and visualize architecture.
From concept generation to material simulation and atmospheric rendering, new AI tools and creative workflows are changing how designers think and produce visual work.

Your Task

Explore one example where AI meets computer graphics — and share your findings with the group.

Choose one focus:

1. A new AI-based tool relevant for visual or spatial design (e.g. diffusion-based rendering, AI-assisted modeling, material generation, procedural workflows, etc.)
2. An artist, designer, or studio who integrates AI into their architectural or visual practice
3. An experimental project or installation that rethinks representation or authorship through AI

Goal

Bring one inspiring case to the group — explain what it does, how it works, and why it matters for the future of architectural visualization and design thinking.

Format

• Short in-class presentation: max. 10 minutes
• Any format is fine — slides, live demo, short video, or image sequence
• Focus less on technical detail, more on concept, impact, and aesthetic
• End with your own 1–2 minute reflection:
  What potential or risk do you see for architecture when AI takes part in the visual process?

Tips

• Think of tools like: ComfyUI, Kaedim, Runway, Stable Diffusion, Luma, ControlNet, PromeAI, Wonder Studio…
• Or artists such as: Refik Anadol, Andrés Reisinger, Charlotte Taylor, Zaha Hadid Virtual, or the “Midjourney architecture” community
• Or experimental workflows like: AI-assisted photogrammetry, text-to-3D pipelines, generative cityscapes, AI-driven environmental storytelling

Purpose

The aim is not to teach the tool — but to raise curiosity, reflect critically, and inspire new visual directions in your own work.

01 Tutorials

Choose a Tutorial (one of the recommended) either from the list or by your own choice. We need to get into “Low Poly” modelling, Scene setup and managing basic Light/Materials setup. So the tutorial should focus on these topics. In case you are a complete Beginner in DCC software please come back to me to find appropriate lessons. 
Step 1: Learn to model: make some Objects, Animals, Landscapes etc. based on the tutorials provided. It can be whatever you like, preferable more than only the “Giraffe”. 
Step 2: Make an image. Take your models and put them into a scene, means we need a Camera, Light and materials (just diffuse color, no textures!, no FX!). Think about composition, mood, style, a Narrative etc. Play with your camera and light setup like a Photographer. Make an image by rendering your scene, and send me…
For getting an impression of your progress please bring your laptop with the current stage of your work to the course. When done you send me your final outcome per mail (3 images, screenshots, file not needed if i have seen it already).

Skills required after the fourth session:

-Navigating in Blender knowing the Basic Interfaces
-Setting up a Scene (Lights and Camera)
-Making basic Materials (Material Qualities without Texturing)
-Understanding the basic Concepts of Low Poly Modeling (Quads, Topology, Loops, Snapping) 
-Scene Organisation (Right Scale, Copy/Instance, Parent/Child, Modifier, Collections
-Rendering a still image in EVEE and CYCLES (differences?)

Links

Download Blender
List of Resources and Tutorials
https://polyhaven.com
https://ambientcg.com/