KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #ai-agents #developer-tools #typescript #codex #cli #claude-code #mcp #rust #deepseek-harness #ai-agent

PoseCap

★ 176 Open GitHub ↗

No description available.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# PoseCap

Real-time markerless motion capture for **Blender**: a webcam streams full-body [SMPL-X](https://github.com/vchoutas/smplx) poses straight onto your character. Built in collaboration with **Corridor Digital**.

Free and [open source](LICENSE).

## See it in action

Point the engine at a video and it streams live SMPL-X poses. Here it drives the standalone skeleton viewer on two of the repo's test clips — source on the left, the streamed pose on the right:

<sub>These clips are the repo's own test fixtures ([tests/fixtures/video/](tests/fixtures/video/SOURCES.json)), run through the real capture pipeline. The viewer draws approximate bone lengths; inside Blender the same stream drives a full SMPL-X character, shown step by step in the [Getting Started guide](doc/guides/getting-started.md).</sub>

## Get started

1. **Download** the Windows installer from the [latest release](https://github.com/CorridorTech/PoseCap/releases/latest) and run it — no administrator rights needed. The open-source installer is intentionally unsigned; read the [SmartScreen and download-verification steps](doc/guides/getting-started.md#1-install-posecap) before running it.
2. **Follow the [Getting Started guide](doc/guides/getting-started.md)** — it takes you from a clean machine to a character moving on your webcam, with a screenshot at every step. About 20 minutes, most of it downloads you can leave running.

You are guided inside Blender too: a **Getting Started checklist** sits at the top of the PoseCap panel and keeps the capture buttons disabled until you are ready, so you can't click into an error.

| Guide | What it covers |
|---|---|
| [Getting started](doc/guides/getting-started.md) | The full walk-through: install → body models → character → live capture |
| [Set up the body models](doc/guides/smplx-model-setup.md) | The one-time SMPL-X download, with your own free research account |
| [Set up a character](doc/guides/character-setup.md) | Bring in a Mixamo or Unreal character and convert it in one click |
| [Live capture](doc/guides/live-capture.md) | Stream from a webcam or a video, and record the motion to keyframes |

## Status

**Early development — first installable preview.** See the [PRD](doc/product/PRD.md) for scope and roadmap.

## What it is

Two components, one pipeline:

- **Blender extension** — panels and operators inside Blender: start/stop the live stream, record mocap to the timeline, capture single poses, manage SMPL-X body models.
- **Engine bridge** — a separate GPU process wrapping the [PEAR](https://github.com/Pixel-Talk/PEAR) pose-estimation model (single image in, SMPL-X parameters out, real-time rates). Streams poses to Blender over a local socket.

## Who it's for

Blender animators who want believable body animation without mocap suits, markers, or external mocap software. Corridor Digital's production artists are the founding users; if you have Blender and a webcam, the MediaPipe Lite backend gives you an account-free CPU path. PEAR remains available for the higher-fidelity NVIDIA workflow.

## Compatibility (current target)

| Component | Supported now |
|---|---|
| OS | Windows 10 / 11 |
| MediaPipe Lite | CPU-first body capture. No GPU required |
| PEAR | NVIDIA/CUDA, driver R570 or newer; qualified on an RTX 3080. RTX 50-series (Blackwell) supported since v1.0.7; GTX 10-series (Pascal) and older are not |
| Blender | 4.2 LTS minimum, 5.x supported |
| Python | 3.11 (Blender bundled for the extension; uv-managed venv for the engine bridge) |
| Camera | Any webcam, including virtual cameras (e.g. Iriun) |

Other PEAR GPU generations remain unqualified unless a release note says otherwise. Linux support for the engine bridge is on the roadmap; macOS is not currently planned.

## How it works

Two processes, joined by explicit contracts:

1. The engine captures webcam frames, finds the person and estimates their full-body pose on the GPU, and streams those poses to Blender over a local connection.
2. The Blender extension consumes the stream on a background thread and applies poses on the main thread as they arrive — without wiping your existing keyframes.
3. Poses apply pelvis-locked: monocular depth estimation cannot recover trustworthy world position, so world translation stays out until a solid software approach lands (camera tracking is the leading candidate — see the roadmap).

Step-by-step diagrams for every flow (live streaming, capture jobs, install) live in [doc/workflows.md](doc/workflows.md); binding structure lives in [ARCHITECTURE.md](ARCHITECTURE.md).

## Body models and licensing

Two separate licenses apply — the plugin's and the body models':

- **PoseCap itself** is free and open source: GPL-3.0 for the Blender extension (required for Blender API linkage), Apache-2.0 for the contracts, core, and engine-bridge libraries — decided in [ADR-0006](doc/adr/0006-license-split-gpl-addon-apache-libraries.md).
- **The SMPL, SMPL-X, and FLAME body models** are licensed by the Max Planck Institute for research (non-commercial) use. They are **never distributed with PoseCap** — not in this repo, not in the installer. Each user registers on the official MPI sites and accepts the license terms personally; the plugin's setup wizard then automates the download using that user's own account credentials. The [illustrated setup guide](doc/guides/smplx-model-setup.md) walks through it.
- **Commercial production use of the body models** requires a commercial license from [Meshcapade](https://meshcapade.com), independent of the plugin itself being free.

## Roadmap

The product roadmap lives in the [PRD](doc/product/PRD.md), the single owner of scope and sequence.

## Project documentation

- [ARCHITECTURE.md](ARCHITECTURE.md) — layers, boundaries, IPC contracts
- [doc/workflows.md](doc/workflows.md) — sequence diagrams and flowcharts per functionality
- [AGENTS.md](AGENTS.md) — operational guide for contributors and coding agents
- [doc/product/PRD.md](doc/product/PRD.md) — product scope, metrics, roadmap
- [doc/reference/](doc/reference/README.md) — the papers and upstream projects this builds on

## Contributing

See [CONTRIBUTING.md](CONTRIBUTING.md) — setup from clone to green tests, the project map, commit conventions, and the rules CI enforces. Short version: most of the codebase (`contracts/`, `core/`) needs no GPU and no Blender to work on; never commit model files or weights; one concern per signed-off commit.

## Acknowledgements

- **Dean / Corridor Digital** — concept and the production use case driving this
- **Alê Alvaro** ([@alexandremendoncaalvaro](https://github.com/alexandremendoncaalvaro)) — the PoseCap rewrite: architecture, implementation, and installers
- **[PEAR](https://wujh2001.github.io/PEAR/)** (Wu et al., IDEA) — the pose-estimation backbone
- **[Fast SAM 3D Body](https://github.com/yangtiming/Fast-SAM-3D-Body)** (Yang et al.) — roadmap engine backend
- **[Meshcapade / MPI SMPL Blender addon](https://github.com/Meshcapade/SMPL_blender_addon)** — the addon lineage the POC started from
- **SMPL-X** body model by MPI for Intelligent Systems

PoseCap

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README