ReasonPix2Pix: Instruction Reasoning Dataset for Advanced Image Editing

TL;DR

ReasonPix2Pix introduces a new dataset and approach to improve image editing models by enhancing their active reasoning capabilities for understanding and executing complex, implicit, and reasoning-based instructions.

Contribution

The paper presents ReasonPix2Pix, a reasoning-attentive instruction editing dataset that enhances models' ability to perform complex, reasoning-based image editing tasks.

Findings

Model fine-tuned on ReasonPix2Pix outperforms others in instructional editing.

Dataset includes reasoning instructions and realistic images with high variance.

Enhanced reasoning improves performance on implicit and explicit instructions.

Abstract

Instruction-based image editing focuses on equipping a generative model with the capacity to adhere to human-written instructions for editing images. Current approaches typically comprehend explicit and specific instructions. However, they often exhibit a deficiency in executing active reasoning capacities required to comprehend instructions that are implicit or insufficiently defined. To enhance active reasoning capabilities and impart intelligence to the editing model, we introduce ReasonPix2Pix, a comprehensive reasoning-attentive instruction editing dataset. The dataset is characterized by 1) reasoning instruction, 2) more realistic images from fine-grained categories, and 3) increased variances between input and edited images. When fine-tuned with our dataset under supervised conditions, the model demonstrates superior performance in instructional editing tasks, independent of whether the tasks require reasoning or not. The code will be available at https://github.com/Jin-Ying/ReasonPix2Pix.