IE-Bench: Advancing the Measurement of Text-Driven Image Editing for Human Perception Alignment

TL;DR

This paper introduces IE-Bench, a new benchmark and assessment method for evaluating text-driven image editing, aligning better with human perception and providing a standardized way to measure editing quality.

Contribution

The paper presents the first IQA dataset and model specifically designed for text-driven image editing, improving evaluation accuracy and human perception alignment.

Findings

IE-QA outperforms previous metrics in subjective alignment

IE-Bench includes 3,010 MOS scores from 25 human subjects

The dataset covers diverse images and editing prompts

Abstract

Recent advances in text-driven image editing have been significant, yet the task of accurately evaluating these edited images continues to pose a considerable challenge. Different from the assessment of text-driven image generation, text-driven image editing is characterized by simultaneously conditioning on both text and a source image. The edited images often retain an intrinsic connection to the original image, which dynamically change with the semantics of the text. However, previous methods tend to solely focus on text-image alignment or have not aligned with human perception. In this work, we introduce the Text-driven Image Editing Benchmark suite (IE-Bench) to enhance the assessment of text-driven edited images. IE-Bench includes a database contains diverse source images, various editing prompts and the corresponding results different editing methods, and total 3,010 Mean Opinion Scores (MOS) provided by 25 human subjects. Furthermore, we introduce IE-QA, a multi-modality source-aware quality assessment method for text-driven image editing. To the best of our knowledge, IE-Bench offers the first IQA dataset and model tailored for text-driven image editing. Extensive experiments demonstrate IE-QA's superior subjective-alignments on the text-driven image editing task compared with previous metrics. We will make all related data and code available to the public.