# Gaussian Splatting Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/gaussian-splatting
> Markdown URL: https://aitinkerers.org/technologies/gaussian-splatting.md
> Technology record last updated: 2026-04-20T23:22:13Z
> Generated: 2026-09-20T15:44:03Z

A novel view synthesis method: it models a scene with millions of explicit 3D Gaussians, enabling photorealistic, real-time radiance field rendering at $\ge 100$ FPS.

Gaussian Splatting (3DGS) is a 2023 breakthrough for 3D reconstruction and novel view synthesis, introduced by Bernhard Kerbl et al.. It replaces the slow, implicit neural network of NeRF with an explicit, high-fidelity representation: a large collection of 3D Gaussians (ellipsoids). Each Gaussian is defined by parameters like position, anisotropic covariance, and color, and is optimized using a fast, differentiable Gaussian rasterizer. This explicit structure delivers massive performance gains: training is up to 50 times faster than NeRF, and rendering achieves true real-time rates, consistently hitting $\ge 100$ FPS at 1080p resolution.

- Official technology site: https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/
- Public AI Tinkerers demos and talks: 5
- Result page: 1 of 1

## Recent Public Talks and Demos

### [Your Webcam Knows the Geometry - Real-Time Relighting & Single-Shot Novel Views](https://lausanne.aitinkerers.org/talks/rsvp_cMMuinG2WV4)

Two single-view inverse-rendering systems that reconstruct scene geometry from one camera and re-render it under conditions never captured — one changes the light, the other changes the viewpoint, both fast enough to ship. The Relighting part decomposes a live webcam frame into geometry + material (normals/albedo/shading) and re-renders it under any lighting in real time (~24 FPS), showcasing a variety of lighting conditions (multiple colored lights, orbiting and rainbow-rotating lights) The Novel View Synthesis builds a 3D representation of the scene using Gaussian Splatting and renders it from a different angle. It currently runs at 5 FPS but we are hoping to get closer to real time soon.

- Event context: AI Tinkerers Lausanne June 2026 Meetup — 2026-06-25 — Lausanne
- Public talk page: https://lausanne.aitinkerers.org/talks/rsvp_cMMuinG2WV4

### [Memory House: Building a Living World from a Single Image with Worldlabs](https://toronto.aitinkerers.org/talks/rsvp_th49b5_Gbgc)

Memory House is an experimental narrative project that uses WorldLabs’ worldbuilding tools and AI models to transform still images into explorable 3D memory spaces. The demo will walk through how environments are stitched, layered with spatial audio, and triggered through AI-generated voice and proximity systems. Each room—built from a single 2D image—becomes a playable fragment of memory, blending reality, dream logic, and procedural storytelling into one seamless experience. This session will cover: Building connected rooms using Gaussian splatting and stitching. Integrating spatial audio and interactive triggers with WorldLabs’ engine. Using AI tools like Veo3 (cinematic video generation) and ElevenLabs (VO) for emergent narrative design. How small design irregularities create the illusion of “distorted memory” and emotional presence. To optimize this workflow, I created World Builder Companion, a custom ChatGPT-powered system that analyzes scene outputs and refines prompts to match WorldLabs’ rendering strengths—balancing material density, lighting, and cinematic tone.

- Event context: AI Tinkerers Toronto - October 2025 Meetup at CIBC | Simplii — 2025-10-30 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_th49b5_Gbgc

### [Spatial Computing for the Masses](https://paris.aitinkerers.org/talks/rsvp_1piGehvTShc)

I will present the Splatrograph API, currently the easiest and most hassle-free way to generate Gaussian Splats, the preferred method for 3D reconstruction. Gaussian Splats also serve as a great input for downstream tasks.

- Event context: AI Tinkerers - Paris Meetup on May 15th — 2025-05-15 — Paris
- Public talk page: https://paris.aitinkerers.org/talks/rsvp_1piGehvTShc

### [Spatial Computing for the Masses](https://milan.aitinkerers.org/talks/rsvp_ifHt2FXkdGA)

I will start by presenting what spatial computing is, tracing its history from photogrammetry to Gaussian splats, and explaining its relationship to AI. Then, I’ll give a demo and a brief technical overview of the Splatrograph API.

- Event context: AI Tinkerers Milan - May 8, 2025 — 2025-05-08 — Milan
- Public talk page: https://milan.aitinkerers.org/talks/rsvp_ifHt2FXkdGA

### [Generación de gemelos digitales con IA](https://bogota.aitinkerers.org/talks/rsvp_UH1ZVndWelg)

En la era actual, la reconstrucción de modelos 3D del mundo real ha avanzado significativamente gracias a dos técnicas innovadoras: NERF (Neural Radiance Fields) y Gaussian Splatting. Estas técnicas, entrenadas mediante redes neuronales, están revolucionando diversos sectores como los videojuegos, el metaverso, la realidad virtual y el sector inmobiliario. Durante esta charla, abordaré los conceptos clave detrás de NERF y Gaussian Splatting, desglosando sus fundamentos y el impacto de sus aplicaciones. Además, compartiré con la audiencia herramientas prácticas y un flujo de trabajo replicable, permitiendo a cualquier persona aplicar estas técnicas en sus propios proyectos.

- Event context: Tercer AI Tinkerers Bogotá — 2024-07-25 — Bogotá
- Public talk page: https://bogota.aitinkerers.org/talks/rsvp_UH1ZVndWelg

## Related Technologies

- [COLMAP](https://aitinkerers.org/technologies/colmap) ([Markdown](https://aitinkerers.org/technologies/colmap.md)) — 2 public demos
- [FastAPI](https://aitinkerers.org/technologies/fastapi) ([Markdown](https://aitinkerers.org/technologies/fastapi.md)) — 181 public demos
- [Ngrok](https://aitinkerers.org/technologies/ngrok) ([Markdown](https://aitinkerers.org/technologies/ngrok.md)) — 3 public demos
- [3D Modeling](https://aitinkerers.org/technologies/3d-modeling) ([Markdown](https://aitinkerers.org/technologies/3d-modeling.md)) — 1 public demo
- [3D Reconstruction](https://aitinkerers.org/technologies/3d-reconstruction) ([Markdown](https://aitinkerers.org/technologies/3d-reconstruction.md)) — 1 public demo
- [Brush](https://aitinkerers.org/technologies/brush) ([Markdown](https://aitinkerers.org/technologies/brush.md)) — 1 public demo
- [Data](https://aitinkerers.org/technologies/data) ([Markdown](https://aitinkerers.org/technologies/data.md)) — 8 public demos
- [Diffusion](https://aitinkerers.org/technologies/diffusion) ([Markdown](https://aitinkerers.org/technologies/diffusion.md)) — 2 public demos
- [DINOv3](https://aitinkerers.org/technologies/dinov3) ([Markdown](https://aitinkerers.org/technologies/dinov3.md)) — 1 public demo
- [DINOv3-ConvNext](https://aitinkerers.org/technologies/dinov3-convnext) ([Markdown](https://aitinkerers.org/technologies/dinov3-convnext.md)) — 1 public demo
- [ElevenLabs](https://aitinkerers.org/technologies/elevenlabs) ([Markdown](https://aitinkerers.org/technologies/elevenlabs.md)) — 42 public demos
- [Inference](https://aitinkerers.org/technologies/inference) ([Markdown](https://aitinkerers.org/technologies/inference.md)) — 9 public demos
- [NeRF](https://aitinkerers.org/technologies/nerf) ([Markdown](https://aitinkerers.org/technologies/nerf.md)) — 2 public demos
- [Neural networks](https://aitinkerers.org/technologies/neural-networks) ([Markdown](https://aitinkerers.org/technologies/neural-networks.md)) — 7 public demos
- [OpenCV](https://aitinkerers.org/technologies/opencv) ([Markdown](https://aitinkerers.org/technologies/opencv.md)) — 26 public demos
- [OpenGL](https://aitinkerers.org/technologies/opengl) ([Markdown](https://aitinkerers.org/technologies/opengl.md)) — 2 public demos
- [PlayCanvas](https://aitinkerers.org/technologies/playcanvas) ([Markdown](https://aitinkerers.org/technologies/playcanvas.md)) — 2 public demos
- [PyTorch](https://aitinkerers.org/technologies/pytorch) ([Markdown](https://aitinkerers.org/technologies/pytorch.md)) — 273 public demos