SLAM-Driven Surrealism

Service Description

Early Bird Registrations are now open for the comprehensive 4-week online workshop, open for beginners and designed for architects, designers, and creatives interested in real-time 3D reconstruction, AI-driven environments, and immersive worldbuilding in Unreal Engine. Participants will be introduced to SLAM concept (Simultaneous Localization and Mapping - no coding or prior experience required), Point Clouds, Gaussian Splatting, and Unreal Engine integration to construct surreal, interactive architectural environments. In addition to Point Clouds workflows, participants will explore a guided 2D-to-3D AI pipeline, where flat AI-generated images are transformed into spatially rich 3D models. These assets will then be incorporated into immersive Unreal Engine scenes, expanding the narrative possibilities of the workshop. Workshop Schedule: Week 1 – June 8, 2025 • Presentation • Introduction to SLAM and other 3D reconstruction and scanning methods, Point Clouds and G-splats • Generating concepts and animations for the world building using AI • Introduction to Unreal Engine - Environment part I Week 2 – June 15, 2025 • 2D to 3D AI architectural assets • Optimizing AI-generated assets for surreal environments • Converting Point Clouds into Gaussian Splatting Week 3 – June 22, 2025 • Real-time rendering of Gaussian Splatting assets in Unreal Engine • Procedural worldbuilding & environmental design Week 4 – July 5, 2025 • Final project presentations • Live Q&A and feedback session Each session includes a 20 minute break midway through. Software Requirements SLAM and 2D-to-3D AI (Setup and various tools provided during the week leading to the first session) Unreal Engine 5 (Free): https://www.unrealengine.com/en-US/download Midjourney (Subscription required): https://www.midjourney.com/ Luma AI (Subscription required): https://lumalabs.ai/ Tripo 3D (Account required) https://www.tripo3d.ai/ System Requirements OS: Windows 10/11 (Preferred) or macOS 10.14+ Graphics: Dedicated GPU (NVIDIA RTX 3060 or higher) Memory: 16 GB RAM (Best: 32 GB or more) Duration: 4 Weeks (1 session per week), 4 Hours per Session Time: 1:00 PM–5:00 PM BST Location: Zoom (Link sent post-registration) Instructor: Chantal Matar Level: Beginner Fee: £450 Standard Registration Capacity: 30 Participants Sessions are recorded. Registration is non-refundable.