# CoEditor++: Instruction-based Visual Editing via Cognitive Reasoning

Source: [CoEditor++: Instruction-based Visual Editing via Cognitive Reasoning](<https://www.sota2.com/research/paper/69ae53f42d603e6de3e429d5>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-10T05:00:05.424Z

Authors: Minheng Ni, Yutao Fan, Zhengyuan Yang, Yeli Shen, Yuxiang Wei, Yaowen Zhang, Lijuan Wang, Lei Zhang, Wangmeng Zuo

Published: 2026-01-31T12:20:46.000Z

## Abstract

Recent advances in large multimodal models (LMMs) have enabled instruction-based image editing, allowing users to modify visual content via natural language descriptions. However, existing approaches often struggle with high-level semantic reasoning and visual consistency, particularly under ambiguous or complex instructions. To address these challenges, we propose CoEditor++, a cognitively structured, training-free framework that decomposes editing into "what to edit" and "how to edit" through two cognitive stages with a reflective self-selection mechanism, enabling robust, fine-grained, and interpretable editing. Built entirely from open-sourced components, CoEditor++ requires no additional training or fine-tuning, ensuring transparency and cross-domain applicability. We evaluate CoEditor++ on SmartEdit, a widely used benchmark for general editing, and AltBear, a privacy and compliance-oriented benchmark. Experimental results show that CoEditor++ achieves state-of-the-art performance in both general editing and responsible editing tasks compared with open-sourced models that require training on specialized editing datasets maintaining significantly higher visual consistency. When compared with closed-source models such as Nano Banana Pro or GPT-4o, CoEditor++ preserves comparable instruction following while still substantially outperforming them in visual consistency. Extensive ablation studies confirm that the effectiveness of CoEditor++ benefits from its structured cognitive design rather than any specific model component. Our findings suggest the potential toward cognitive-centric instruction-based image editing.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Instruction-based Image Editing on SmartEdit Understanding](<https://www.sota2.com/research/sota/instruction-based-image-editing-on-smartedit-understanding.md>): Instruction-based Image Editing; SmartEdit Understanding; PSNR: 35.89.

- [Instruction-based Image Editing on SmartEdit Reasoning](<https://www.sota2.com/research/sota/instruction-based-image-editing-on-smartedit-reasoning.md>): Instruction-based Image Editing; SmartEdit Reasoning; PSNR: 41.061.

- [Instruction-based Image Editing on AltBear Responsible](<https://www.sota2.com/research/sota/instruction-based-image-editing-on-altbear-responsible.md>): Instruction-based Image Editing; AltBear Responsible; PSNR: 12.008.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.05518>)

[PDF](<https://arxiv.org/pdf/2603.05518.pdf>)
