Publication
LEDITS++: Limitless Image Editing using Text-to-Image Models
Manuel Brack; Felix Friedrich; Katharina Kornmeier; Linoy Tsaban; Patrick Schramowski; Kristian Kersting; Apolinário Passos
In: Computing Research Repository eprint Journal (CoRR), Vol. abs/2311.16711, Pages 1-21, arXiv, 2023.
Abstract
Text-to-image diffusion models have recently received
increasing interest for their astonishing ability to produce
high-fidelity images from solely text inputs. Subsequent
research efforts aim to exploit and apply their capabili-
ties to real image editing. However, existing image-to-
image methods are often inefficient, imprecise, and of lim-
ited versatility. They either require time-consuming fine-
tuning, deviate unnecessarily strongly from the input im-
age, and/or lack support for multiple, simultaneous edits.
To address these issues, we introduce LEDITS++, an effi-
cient yet versatile and precise textual image manipulation
technique. LEDITS++’s novel inversion approach requires
no tuning nor optimization and produces high-fidelity re-
sults with a few diffusion steps. Second, our methodology
supports multiple simultaneous edits and is architecture-
agnostic. Third, we use a novel implicit masking technique
that limits changes to relevant image regions. We propose
the novel TEdBench++ benchmark as part of our exhaus-
tive evaluation. Our results demonstrate the capabilities of
LEDITS++ and its improvements over previous methods.