🚀 etcGrab Cube — My Browser-Based VR Full-Body Tracking System GitHub: https://github.com/Sami9889/hand-grab-cube
I’ve been working on a project called etcGrab Cube, a fully modular VR-ready system for full-body tracking, hand tracking, face tracking, ragdoll physics, and advanced avatar animation — all running directly in the browser with no extra hardware.
The system combines MediaPipe Pose, Hands, and FaceMesh with three.js to deliver real-time 3D motion capture, expressive avatars, and interactive VR experiences. My goal with this project is to push what’s possible using only a camera and the web.
⸻
🌟 What etcGrab Cube Can Do
🔥 Full-Body Tracking Captures your movement through MediaPipe Pose, supports world landmarks, and stay…
🚀 etcGrab Cube — My Browser-Based VR Full-Body Tracking System GitHub: https://github.com/Sami9889/hand-grab-cube
I’ve been working on a project called etcGrab Cube, a fully modular VR-ready system for full-body tracking, hand tracking, face tracking, ragdoll physics, and advanced avatar animation — all running directly in the browser with no extra hardware.
The system combines MediaPipe Pose, Hands, and FaceMesh with three.js to deliver real-time 3D motion capture, expressive avatars, and interactive VR experiences. My goal with this project is to push what’s possible using only a camera and the web.
⸻
🌟 What etcGrab Cube Can Do
🔥 Full-Body Tracking Captures your movement through MediaPipe Pose, supports world landmarks, and stays stable during fast motion, jumping, or partial visibility.
🖐 Hand & Finger Tracking + Gesture Detection Tracks fingers, pinch/grab gestures, double pinches, and presence/loss of hands. Perfect for interactions, UI control, and gameplay systems.
🙂 Face Tracking FaceMesh adds facial landmark tracking for expressions, avatar reactions, and more.
🎥 Multi-Camera Fusion Supports multiple webcams at once for improved 3D accuracy, depth estimation, and a better understanding of body position in space.
🧍 Modular Advanced Avatar Includes a detailed avatar with expressive motion, accessories, animation layers, outlines, smoothing, and modular components you can swap or extend.
🧩 Ragdoll Physics Toggle ragdoll mode anytime. The system smoothly blends between physics simulation and tracked movement for realistic falling and interaction.
🎮 WebXR VR Support Enter VR directly in the browser. Supports VR controllers, tracked hands (if available), and haptics for interaction feedback.
🛠 Full UI Controls Everything is adjustable: • Ragdoll • VR mode • HUD overlays • Tracking mode • Smoothing • Low-perf mode • Multi-camera selection • Snapshot capture
🪄 Gesture & Interaction Events High-level events fire for pinches, releases, hand presence, double-pinches, and more, making integration with apps, games, and tools seamless.
📸 Snapshot & Debug Tools Capture the 3D scene, view tracking overlays, and debug pose/hand/face landmarks.
🎞 Camera Fallback Mode Use a test video input when no real camera is available.
🔒 Robust Error Handling Handles camera permissions, disconnects, and performance gracefully for a smooth experience.
⸻
🎯 Why I Built It
I wanted a lightweight, browser-based full-body tracking system that: • Doesn’t require VR trackers • Doesn’t need special gloves or sensors • Works with a laptop webcam or multiple cameras • Supports VR and physics • Can become a base for future games, tools, and creative experiments
etcGrab Cube is my solution — a flexible foundation for VR, animation, motion capture, and interactive 3D.
⸻
🔗 GitHub Repository
https://github.com/Sami9889/hand-grab-cube
If you’re into VR, tracking, or web-based 3D, feel free to check it out, fork it, or reach out with ideas!