Skip to main content

Envisioning is an emerging technology research institute and advisory.

LinkedInInstagramGitHub

2011 — 2026

research
  • Observatory
  • Newsletter
  • Methodology
  • Origins
  • Vocab
services
  • Signals Session
  • Bespoke Projects
  • Build Sessions
  • Use Cases
  • Readinessfree
  • Signals
  • Free scan↗free
impact
  • ANBIMAFuture of Brazilian Capital Markets
  • IEEECharting the Energy Transition
  • Horizon 2045Future of Human and Planetary Security
  • WKOTechnology Scanning for Austria
solutions
  • Innovation
  • Strategy
  • Consultants
  • Foresight
  • Associations
  • Governments
resources
  • Partners
  • Coding for Non-Coders
  • How We Work
  • Data Visualization
  • Multi-Model Method
  • FAQ
  • Security & Privacy
about
  • Manifesto
  • Community
  • Events
  • Support
  • Contact
ResearchServicesSignalsAbout
ResearchServicesSignalsAbout
  1. Home
  2. Research
  3. Prism
  4. 3D Gaussian Splatting Engines

3D Gaussian Splatting Engines

Real-time rendering of photorealistic 3D scenes from multi-view photos using GPU-rasterized Gaussian splats
Back to PrismView interactive version

3D Gaussian splatting engines represent scenes as millions of anisotropic Gaussian blobs with per-blob color, opacity, and covariance. Instead of ray marching through neural networks, they rasterize these splats directly on the GPU, achieving NeRF-level realism at interactive frame rates and allowing editors to edit splats like particles. Capture pipelines convert multi-view images into splat clouds in minutes, enabling portable files that stream efficiently to WebGL, Unreal, or custom viewers.

Studios use splats to build volumetric location libraries for AR scouting, lightweight digital twins for sports replays, and immersive tourism experiences that run on phones. Because the format supports progressive refinement, streaming services can send low-res previews first, then layer in detail as bandwidth allows. Paired with LiDAR data, splats fill gaps with photoreal draw-ins, helping virtual production teams align CG and practical sets.

Ecosystem maturity (TRL 5) hinges on tooling: DCC apps need splat-aware editing, and compression standards must emerge to keep files manageable. Open-source viewers from Graphdeco and Luma Labs demonstrate what’s possible, and standards bodies discuss adding splat modes to glTF. Expect Gaussian splatting to sit alongside NeRFs as a pragmatic option where real-time playback and editability trump absolute compression rates.

TRL
5/9Validated
Impact
4/5
Investment
3/5
Category
Software

Related Organizations

INRIA logo
INRIA

France · Research Lab

100%

The French National Institute for Research in Digital Science and Technology, heavily involved in AI research and Scikit-learn.

Researcher
Luma AI logo
Luma AI

United States · Startup

98%

Creators of Dream Machine, a high-quality video generation model, and 3D capture technology.

Developer
Polycam logo
Polycam

United States · Startup

95%

A leading 3D capture application for mobile devices.

Developer
Postshot

United States · Startup

92%

Software dedicated to creating and training Radiance Fields and Gaussian Splats.

Developer
Max Planck Institute for Informatics logo
Max Planck Institute for Informatics

Germany · Research Lab

90%

Pioneers in Neural Radiance Fields (NeRF) and light-field reconstruction algorithms.

Researcher
Niantic logo
Niantic

United States · Company

90%

AR platform company that develops the Lightship ARDK and owns Scaniverse, a 3D scanning app leveraging LiDAR.

Deployer
PlayCanvas logo
PlayCanvas

United Kingdom · Company

88%

A WebGL game engine owned by Snap Inc.

Developer
NVIDIA logo
NVIDIA

United States · Company

85%

Developing foundation models for robotics (Project GR00T) and vision-language models like VILA.

Researcher
Spline logo
Spline

United States · Startup

85%

A web-based 3D design tool enabling real-time collaboration.

Deployer
Unity Technologies logo
Unity Technologies

United States · Company

80%

Provides the High Definition Render Pipeline (HDRP) which supports real-time ray tracing for gaming and industrial visualization.

Developer

Supporting Evidence

Evidence data is not available for this technology yet.

Research this in Signals

Scan 3D Gaussian Splatting Engines for yourself.

Signals turns a topic into a sourced research record you can inspect and rerun. Your first scan is free, and this one starts with 3D Gaussian Splatting Engines already loaded, so edit it or scan as is.