KV Inversion: KV Embeddings Learning for Text-Conditioned Real Image Action Editing

The study introduces KV Inversion, a novel method for text-conditioned real image action editing. The method addresses the problem of action editing, ensuring the edited results match the corresponding action, while retaining the texture and identity of the original image. The method doesn’t require training the Stable Diffusion model or scanning a large-scale dataset. The potential applications of this technology include comic book production, video editing, and advertising material production.

Publication date: 28 Sep 2023
Project Page: https://arxiv.org/abs/2309.16608v1
Paper: https://arxiv.org/pdf/2309.16608