Luma AI - multiple video shots from a single NeRF capture

The advent of NeRFs is set to revolutionize video production, empowering directors with the freedom to manipulate camera angles and motion in post production, unrestricted by the shots captured during filming. After gaining access to the Luma AI beta program, one of the first things I scanned was my motorcycle. Although the online editor was not yet available, the NeRF data for my scene was saved on the Luma server. Several months later, I added the racer character to the scene and captured different shots in the same location. This is a brief rundown of the steps involved in this process: 1. Utilize Luma AI app to scan the location. 2. Download the 3D model and open it in Blender. 3. Animate the Blender camera and export it as an FBX file. 4. Import the motion paths to the Luma AI online editor. 5. Render the full NeRF scene videos. 6. Render the motorcycle over a green background. 7. Render the character with a transparent background. 8. Composite the layers and render the