Upload README.md with huggingface_hub

README.md

@@ -55,6 +55,94 @@ pipeline_tag: reinforcement-learning
 - Injured walking
 - Boxing stance
 
+## VR Whole-Body Teleoperation
+
+SONIC supports real-time whole-body teleoperation via PICO VR headset, enabling natural human-to-robot motion transfer for data collection and interactive control.
+
+<div align="center">
+<table>
+<tr>
+<td align="center"><b>Walking</b></td>
+<td align="center"><b>Running</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/teleop_walking.gif" width="400"></td>
+<td align="center"><img src="media/teleop_running.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Sideways Movement</b></td>
+<td align="center"><b>Kneeling</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/teleop_sideways.gif" width="400"></td>
+<td align="center"><img src="media/teleop_kneeling.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Getting Up</b></td>
+<td align="center"><b>Jumping</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/teleop_getup.gif" width="400"></td>
+<td align="center"><img src="media/teleop_jumping.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Bimanual Manipulation</b></td>
+<td align="center"><b>Object Hand-off</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/teleop_bimanual.gif" width="400"></td>
+<td align="center"><img src="media/teleop_switch_hands.gif" width="400"></td>
+</tr>
+</table>
+</div>
+
+## Kinematic Planner
+
+SONIC includes a kinematic planner for real-time locomotion generation — choose a movement style, steer with keyboard/gamepad, and adjust speed and height on the fly.
+
+<div align="center">
+<table>
+<tr>
+<td align="center" colspan="2"><b>In-the-Wild Navigation</b></td>
+</tr>
+<tr>
+<td align="center" colspan="2"><img src="media/planner/planner_in_the_wild_navigation.gif" width="800"></td>
+</tr>
+<tr>
+<td align="center"><b>Run</b></td>
+<td align="center"><b>Happy</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/planner/planner_run.gif" width="400"></td>
+<td align="center"><img src="media/planner/planner_happy.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Stealth</b></td>
+<td align="center"><b>Injured</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/planner/planner_stealth.gif" width="400"></td>
+<td align="center"><img src="media/planner/planner_injured.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Kneeling</b></td>
+<td align="center"><b>Hand Crawling</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/planner/planner_kneeling.gif" width="400"></td>
+<td align="center"><img src="media/planner/planner_hand_crawling.gif" width="400"></td>
+</tr>
+<tr>
+<td align="center"><b>Elbow Crawling</b></td>
+<td align="center"><b>Boxing</b></td>
+</tr>
+<tr>
+<td align="center"><img src="media/planner/planner_elbow_crawling.gif" width="400"></td>
+<td align="center"><img src="media/planner/planner_boxing.gif" width="400"></td>
+</tr>
+</table>
+</div>
+
 ## Quick Start
 
 📚 **See the [Quick Start Guide](https://zhengyiluo.github.io/GR00T-WholeBodyControl/getting_started/quickstart.html)** for step-by-step instructions on:
@@ -71,85 +159,9 @@ pipeline_tag: reinforcement-learning
 
 ## Model Checkpoints
 
-Pre-trained SONIC policy checkpoints are included in the repository under `gear_sonic_deploy/models/`. The models are optimized for:
-- Real-time inference on NVIDIA Jetson platforms
-- Low-latency control (< 50ms)
-- Robust sim-to-real transfer
-
-## Architecture
-
-SONIC uses a transformer-based policy architecture trained with:
-- **Training Framework**: Isaac Lab 2.3.0
-- **Backend**: PyTorch
-- **Deployment**: C++ inference with TorchScript
-- **Control Frequency**: 20-50 Hz
-- **Observation Space**: Proprioceptive sensing + motion tracking targets
-- **Action Space**: Joint position/velocity targets
-
-## Training Data
-
-SONIC is trained on large-scale human motion capture datasets including:
-- Locomotion sequences (walking, running, jumping)
-- Full-body manipulation tasks
-- Multi-modal movement styles
-- Ground-up recovery behaviors
-
-The model learns through motion tracking, which provides a scalable supervisory signal for acquiring diverse motor skills.
-
-## VR Teleoperation
-
-SONIC supports real-time whole-body teleoperation for data collection and interactive control:
-
-- **Hardware**: PICO VR headset + controllers
-- **Tracking**: Full-body motion retargeting
-- **Latency**: < 100ms end-to-end
-- **Applications**: Demonstration collection, direct control, behavior annotation
-
-See the [VR Teleoperation Guide](https://zhengyiluo.github.io/GR00T-WholeBodyControl/tutorials/vr_wholebody_teleop.html) for setup instructions.
-
-## Intended Use
-
-### Primary Use Cases
-
-✅ **Research & Development**
-- Humanoid robot control research
-- Whole-body motion learning
-- Sim-to-real transfer studies
-- Teleoperation system development
-
-✅ **Robotics Applications**
-- Mobile manipulation tasks
-- Human-robot interaction
-- Demonstration collection
-- Behavior cloning baselines
-
-### Out-of-Scope Use
-
-❌ This model is **NOT** intended for:
-- Safety-critical applications without additional validation
-- Autonomous operation in uncontrolled environments
-- Direct deployment without proper safety measures
-- Applications requiring certified safety guarantees
-
-## Limitations
-
-- **Hardware Specific**: Models are tuned for specific humanoid platforms (primarily Unitree G1)
-- **Sim-to-Real Gap**: Performance may vary across different hardware configurations
-- **Safety**: Requires proper safety protocols during deployment
-- **Generalization**: Best performance on behaviors similar to training distribution
-- **Environment**: Designed for structured indoor environments
-
-## Hardware Requirements
-
-### Deployment Platform
-- **Recommended**: NVIDIA Jetson Orin (32GB or 64GB)
-- **Minimum**: NVIDIA GPU with CUDA support
-- **Robot**: Compatible humanoid platforms (Unitree G1, H1, etc.)
+Pre-trained SONIC policy checkpoints (ONNX format) are available in the `policy/release/` directory of this repository. The models are optimized for real-time inference on NVIDIA Jetson platforms with low-latency control and robust sim-to-real transfer.
 
-### Development Platform
-- **GPU**: NVIDIA GPU with 16GB+ VRAM (for training)
-- **RAM**: 32GB+ recommended
-- **Storage**: 100GB+ for datasets and checkpoints
+See the [Installation Guide](https://zhengyiluo.github.io/GR00T-WholeBodyControl/getting_started/installation_deploy.html) for instructions on downloading and using the model checkpoints.
 
 ## Documentation