Kernekoncepter
FastVideoEdit proposes an efficient zero-shot video editing approach inspired by Consistency Models, reducing editing time while maintaining high quality.
Statistik
Diffusion models have demonstrated remarkable capabilities in text-to-image and text-to-video generation.
FastVideoEdit eliminates the need for time-consuming inversion or additional condition extraction.
Experimental results validate the state-of-the-art performance and speed advantages of FastVideoEdit.
Citater
"FastVideoEdit offers efficient, consistent, high-quality, and text-aligned editing capabilities for both artificial and natural videos."
"Our method enables direct mapping from source video to target video with strong preservation ability utilizing a special variance schedule."