CROP: Expert-Aligned Image Cropping via Compositional Reasoning and Optimizing Preference

2026-05-09

AI Summary

Existing image cropping methods struggle to balance compositional aesthetics and expert aesthetic preferences in complex scenes, often lacking deep understanding of compositional principles due to their reliance on saliency prediction or retrieval mechanisms. This work reframes aesthetic cropping as a multimodal reasoning task and introduces a three-stage “analyze–propose–decide” framework grounded in vision-language models, emulating the stepwise reasoning process of professional photographers. To further align cropping decisions with human expertise, the method incorporates an expert preference alignment mechanism. It achieves the first interpretable modeling of compositional aesthetics, significantly improving both the aesthetic quality of cropped images and their consistency with expert judgments across multiple benchmark datasets.

Abstract

Aesthetic image cropping aims to enhance the aesthetic quality of an image by improving its composition through spatial cropping. Previous methods often rely on saliency prediction or retrieval augmentation, ignoring the task's core requirement: a deep understanding of composition and aesthetics. Consequently, saliency-based methods struggle to make compositional trade-offs in complex scenes, while retrieval-based methods blindly refer to similar cases, lacking adaptive reasoning for unique scenes. Both approaches fail to align their automated cropping results with those of human experts. To address the above issues, we propose a novel paradigm that reformulates aesthetic cropping as a multimodal reasoning task, aiming to activate the VLM's analytical and comprehension capabilities in aesthetics. We design a Compositional Reasoning and Optimizing Preference method (CROP) that directs the VLM to think like a professional photographer. It deconstructs a complex and subjective aesthetic problem into an "analysis-proposal-decision" process, reasoning step by step through the analysis of scene elements and compositional principles. Meanwhile, our expert preference alignment module makes the model's decision consistent with human expert aesthetics. Extensive experiments across multiple datasets validate our method's superiority and component effectiveness.

Problem

• aesthetic image cropping
• composition
• expert alignment
• visual aesthetics
• image composition

Innovation

• compositional reasoning
• expert preference alignment
• aesthetic image cropping
• vision-language model
• multimodal reasoning