This is an advance, hands on, 2 days course. We will learn key foundational concepts between computer graphics and machine learning that will allow us to bridge these two words with practical examples and hands-on activities.

Patricio González Vivo is an artist with an experimental tinkering-based practice inspired by the mechanics and aesthetics of mapping instruments, including star maps, compasses, telescopes, and satellites. His work is generative and interactive; he often creates symbolic elements to invite people into an experience of presence, play, and imagination. As an art therapist, he witnessed trauma healed through play with symbolic substitutes. His art uses symbols to open the possibility of what-if, to free us from the gravity of the factual and allow us to re-imagine.

Two full days, one as an instructed tutorial workshop, the other as self-guided project work.

Day 1

Morning: From zero to hero on Computer Graphics

• Course overview
• CG Raster pipeline recap
• CG primer on some key concept using Blender
• Deep dive on the same concepts but using GlslViewer (GLSL code) + LYGIA

Afternoon: Machine Learning Foundation

• ML primer / setup / Intro
• Using models
• Try the models on your own assets (pictures)
• Back-feeding it to a rendering engine: Blender / GlslViewer
• Workshop

Day 2

Morning: Bringing it all together

• Projects sharing
• Presentation of:
• Computational Photography (Forensic Architecture)
• Interactivity
• Video: how this works with video
• Web: how this can be bring
• Advance OpenGL / GLSL techniques using GlslViewer for real time editing
• Bokeh effects
• Point clouds (GPGPU)
• PingPong (doubleBuffer)

Afternoon: Workshop

• Workshop time
• Sharing of projects
• Aperitivo + Dinner

Basic knowledge shader coding (GLSL/HLSL/Metal or WGSL) - if you don’t know if you qualify please complete up to chapter 13 in The Book Of Shaders".
Basic knowledge on python programming.
Software: Blender (Addons for OSC, fspy, colmap, Git client, Ffmpeg, VS Code, glslViewer (compiled and installed from source: linux, macos, windows), conda installed + Python + torch.
Hardware: a computer capable of running pytorch (RTX 30XX >=).