Command Palette

Search for a command to run...

UnylyUnyly
Browse all

Lumina

FreeNot checked

Lumina: An AI-powered Active Capture system for high-fidelity 3D Head Avatars. Uses real-time pose guidance and 3D Gaussian Splatting to transform standard webc

GitHubEmbed

About

Lumina: An AI-powered Active Capture system for high-fidelity 3D Head Avatars. Uses real-time pose guidance and 3D Gaussian Splatting to transform standard webcam feeds into animatable digital twins.

README

Lumina logo

Lumina

Lumina is a desktop-first active-capture system for fast 3D head reconstruction.
V1 is focused on a practical local workflow: capture a single neutral head from a laptop webcam, fit a coarse full-head model with rough hair coverage, and export a GLB on macOS.

V1 Goal

  • Single-person only
  • Neutral head only
  • Laptop webcam only
  • Full head with rough hair
  • GLB export only
  • Local reconstruction on macOS
  • Target turnaround under 5 minutes

Current Architecture

  • apps/director-desktop Electron shell for local capture, session control, and desktop runtime integration
  • apps/director-web React + Vite + Three.js renderer embedded inside Electron
  • services/recon-api Optional local orchestration service for manifests, pipeline control, and future decoupling
  • services/recon-worker Local reconstruction worker for FLAME fitting and GLB export
  • packages/protocol Shared TypeScript protocol and event contracts

Product Direction

Lumina v1 is intentionally not chasing maximum fidelity. The current target is a robust and fast local capture loop that:

  • guides the user through a fixed neutral pose sequence
  • auto-accepts stable frames
  • stores a local JSON session manifest
  • prepares a local reconstruction handoff

The long-term direction can still include higher-end Gaussian Splatting pipelines, but that is not the default production path for the first macOS release.

Local Development

Requirements:

  • Node.js 20+
  • pnpm
  • Python 3.11+
  • uv
  • macOS with webcam access

Install and run:

corepack enable
corepack prepare [email protected] --activate
pnpm install
uv sync
pnpm dev

Important pnpm Note

On some machines, pnpm blocks install/build scripts by default.
If Electron fails with Electron failed to install correctly, approve build scripts and rebuild:

pnpm approve-builds
pnpm rebuild
pnpm --filter director-desktop exec electron --version
pnpm dev

Approve at least:

  • electron
  • esbuild

Repository Notes

  • Logo asset: apps/director-web/public/assets/lumina-logo.jpeg
  • Local capture sessions are written under ~/Documents/LuminaSessions
  • The desktop app currently includes:
    • webcam preview
    • pose sequence state
    • automatic frame acceptance scaffold
    • manual retry
    • local manifest writes through Electron IPC

Status

This repository currently contains the desktop-first scaffold, capture-loop foundation, local IPC bridge, and reconstruction boundaries for Lumina v1.

from github.com/oemercemtabar/Lumina

Install Lumina in Claude Desktop, Claude Code & Cursor

Recommended · one command, every IDE
unyly install lumina

Installs into Claude Desktop, Claude Code, Cursor & VS Code — handles npx, uvx and build-from-source repos for you.

First time? Get the CLI: curl -fsSL https://unyly.org/install | sh

Or configure manually

Run in your terminal:

claude mcp add lumina -- uvx lumina

Step-by-step: how to install Lumina

FAQ

Is Lumina MCP free?

Yes, Lumina MCP is free — one-click install via Unyly at no cost.

Does Lumina need an API key?

No, Lumina runs without API keys or environment variables.

Is Lumina hosted or self-hosted?

Self-hosted: the server runs locally on your machine via the install command above.

How do I install Lumina in Claude Desktop, Claude Code or Cursor?

Open Lumina on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.

Related MCPs

Compare Lumina with

Not sure what to pick?

Find your stack in 60 seconds

Author?

Embed badge for your README

Browse similar

All ai MCPs