Back to directory

SAM3DBody-cpp

A standalone C++ inference engine for real-time, multi-person 3D body pose reconstruction from a single camera, exporting motion-capture BVH files or MPEG avatar containers with no Python dependency at runtime.

SAM3DBody-cpp

SAM3DBody-cpp is a standalone C++ inference engine for SAM-3D-Body, a model that reconstructs 3D human body pose from a single camera image with no Python dependency at runtime. Given an image, it produces per-person body pose parameters, camera translation, and optionally a full 3D mesh with seventy body and hand keypoints, using ONNX Runtime and ggml for inference.

For motion capture, it can export a standard BVH file per detected person, keeping each person's identity stable across frames with a bounding-box tracker and auto-resizing each skeleton's bone lengths to the person it followed, ready to load into Blender or another animation tool. It can also export an MPEG Avatar Representation Format container per person, carrying the mesh and skin weights along with the skeleton.

It also ships Python bindings that call the compiled library through ctypes, and pre-built ONNX, GGUF, and LBS model files are hosted separately for CPU, CUDA, or TensorRT use.

More in Inference

See all 770
  • OpenDecision

    deepanwadhwa.github.io

    An open-source Python package that answers typed questions about application state or documents using a local natural-language-inference model, returning structured choice, score, yes/no or evidence-relation values instead of free text.

  • Lifeboat

    iterate.ai

    A local inference server that puts an OpenAI- and Anthropic-compatible API in front of open-weight language models running on hardware you control, from a laptop to a GPU cluster.

  • Claude S40

    github.com

    An unofficial Claude chat client for 2007-era Nokia Series 40 phones, built as a Java ME app plus a small Go relay server, with web search, reading mode and saved chats.