SpaceEdit: Learning a Unified Editing Space for Open-Domain Image Editing

TL;DR

This paper introduces SpaceEdit, a unified, semantic editing space for open-domain image editing that enables intuitive manipulation and supports diverse downstream tasks like language-guided editing and style clustering.

Contribution

The paper proposes a novel learned editing space that is more semantic and manipulable than traditional operation-based spaces, enabling versatile image editing applications.

Findings

Superior performance on language-guided editing tasks

Effective in style clustering and retrieval

Supports multimodal and personalized editing

Abstract

Recently, large pretrained models (e.g., BERT, StyleGAN, CLIP) have shown great knowledge transfer and generalization capability on various downstream tasks within their domains. Inspired by these efforts, in this paper we propose a unified model for open-domain image editing focusing on color and tone adjustment of open-domain images while keeping their original content and structure. Our model learns a unified editing space that is more semantic, intuitive, and easy to manipulate than the operation space (e.g., contrast, brightness, color curve) used in many existing photo editing softwares. Our model belongs to the image-to-image translation framework which consists of an image encoder and decoder, and is trained on pairs of before- and after-images to produce multimodal outputs. We show that by inverting image pairs into latent codes of the learned editing space, our model can be leveraged for various downstream editing tasks such as language-guided image editing, personalized editing, editing-style clustering, retrieval, etc. We extensively study the unique properties of the editing space in experiments and demonstrate superior performance on the aforementioned tasks.