Hands-off Image Editing: Language-guided Editing without any Task-specific Labeling, Masking or even Training

TL;DR

This paper introduces a novel image editing method guided solely by language instructions, eliminating the need for task-specific labels, masks, or training, and demonstrating competitive performance.

Contribution

It presents a task-supervision-free approach to language-guided image editing, enabling more scalable and adaptable editing without specialized training.

Findings

Achieves competitive image editing results without task-specific supervision

Eliminates the need for labels, masks, or training in instruction-guided editing

Demonstrates high effectiveness in diverse editing scenarios

Abstract

Instruction-guided image editing consists in taking an image and an instruction and deliverring that image altered according to that instruction. State-of-the-art approaches to this task suffer from the typical scaling up and domain adaptation hindrances related to supervision as they eventually resort to some kind of task-specific labelling, masking or training. We propose a novel approach that does without any such task-specific supervision and offers thus a better potential for improvement. Its assessment demonstrates that it is highly effective, achieving very competitive performance.