A method for editing NeRF scenes using text-instructions is proposed.
The method uses an image-conditioned diffusion model to optimize the 3D scene based on the edit instruction.
NeRF: Neural Radiance Fields, a technique for synthesizing novel views of a scene
Text-instructions: Written directions used to guide the editing process
Image-conditioned diffusion model: A model that iteratively edits input images while optimizing the underlying scene
InstructPix2Pix: The image-conditioned diffusion model used in this method
Optimized 3D scene: The final edited scene that respects the provided edit instruction