🦅 OpenBird-artnfull AI Models & Motion Datasets

Official AI weights, 15-stage Physical AI trajectory datasets, and simulation preview for the OpenBird-artnfull Bipedal VTOL Robot Platform.

---

🌟 Overview

This repository hosts pre-trained policy checkpoints, 3D MuJoCo assets, and multi-modal trajectory datasets for OpenBird-artnfull, an avian-inspired Physical AI companion robot featuring:

• Digitigrade Bipedal Locomotion (Avian joint kinematics & dynamic walking)
• Vertical Stowed & Deployed Duct Wings (180° horizontal VTOL deployment)
• Look-Down Eye-Gaze & Beak Pitch Control (Independent 60° downward ground observation)
• Stereo Eye Cameras + Head FPV Sensory Suite

---

📂 Repository Contents

2_huggingface_public/
├── policies/               # Trained neural network weights (.pt, .onnx)
│   └── baseline_flight/    # Baseline VTOL takeoff, hover, and landing models
├── datasets/               # MuJoCo / LeRobot compatible motion trajectories
│   └── demonstration/      # Teleoperation & autonomous behavioral data
├── outputs/eval/           # Official evaluation metrics & trajectory logs
│   ├── eval_info.json      # Rollout evaluation statistics
│   └── videos/             # High-definition evaluation rollouts
├── robot_bird/             # MuJoCo 3D kinematics, FSM controller & XML models
└── README.md               # Model & dataset card

---

🚀 Quick Usage with Python

# Clone simulation environment from GitHub
# git clone https://github.com/artnfull-bot/OpenBird-artnfull.git

# Download policies & datasets using huggingface_hub
from huggingface_hub import snapshot_download

model_path = snapshot_download(repo_id="artnfull/OpenBird-artnfull", repo_type="model")
print(f"Downloaded assets to: {model_path}")