Overview

The Video to SimReady workflow creates a simulation-ready 3D asset from a short video clip. By analyzing frames across the video, the system extracts multiple viewpoints of an object and uses multi-view 3D reconstruction techniques to build a complete mesh.

<aside> 🔬 ⚠️ This feature is still in development and may have limitations. Results may vary depending on video quality, lighting consistency, and object motion. For production use, Image to SimReady is recommended.

</aside>

How It Works

1. Frame Extraction — Key frames are automatically extracted from the video at regular intervals or at points of maximum viewpoint diversity.
2. Multi-View Reconstruction — The extracted frames are processed as a multi-view input set, using structure-from-motion and neural reconstruction techniques to infer 3D geometry.
3. Mesh and Texture Generation — The reconstructed 3D geometry is meshed, cleaned, and textured using the same pipeline as Image to SimReady.
4. SimReady Pipeline — The asset is then processed through the physics and collision generation stage.

Best Practices for Input Video

• Keep the object centered in the frame throughout the video.
• Move slowly around the object to capture all angles — avoid fast motion or abrupt cuts.
• Use consistent, even lighting — avoid shadows or reflections that obscure the object surface.
• Use a plain, contrasting background.
• Keep the object stationary — do not rotate or move the object itself during recording.
• Recommended duration: 10–30 seconds for complete 360° coverage.

Learn more in the Physics and Collision Generation documentation.