MoEdit: On Learning Quantity Perception for Multi-object Image Editing

2025-03-13

AI Summary

Existing multi-object image editing methods struggle to simultaneously achieve fine-grained control over individual object attributes and maintain global consistency in object count, spatial distribution, and semantic coherence—leading to distortions in quantity, layout, and semantics post-editing. To address this, we propose MoEdit, the first framework integrating Feature-decoupled Compensation (FeCom) and Unsupervised Quantity-aware Token Attention (QTTN) into the Stable Diffusion architecture. FeCom enhances object-level attribute disentanglement, while QTTN enables count-consistent modeling without additional annotations or auxiliary models. MoEdit uniformly supports high-fidelity style transfer, object re-generation, and background reconstruction. Evaluated on multi-object editing benchmarks, it achieves state-of-the-art performance, significantly improving joint consistency across three critical dimensions: numerical accuracy, spatial plausibility, and visual fidelity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Multi-object images are prevalent in various real-world scenarios, including augmented reality, advertisement design, and medical imaging. Efficient and precise editing of these images is critical for these applications. With the advent of Stable Diffusion (SD), high-quality image generation and editing have entered a new era. However, existing methods often struggle to consider each object both individually and part of the whole image editing, both of which are crucial for ensuring consistent quantity perception, resulting in suboptimal perceptual performance. To address these challenges, we propose MoEdit, an auxiliary-free multi-object image editing framework. MoEdit facilitates high-quality multi-object image editing in terms of style transfer, object reinvention, and background regeneration, while ensuring consistent quantity perception between inputs and outputs, even with a large number of objects. To achieve this, we introduce the Feature Compensation (FeCom) module, which ensures the distinction and separability of each object attribute by minimizing the in-between interlacing. Additionally, we present the Quantity Attention (QTTN) module, which perceives and preserves quantity consistency by effective control in editing, without relying on auxiliary tools. By leveraging the SD model, MoEdit enables customized preservation and modification of specific concepts in inputs with high quality. Experimental results demonstrate that our MoEdit achieves State-Of-The-Art (SOTA) performance in multi-object image editing. Data and codes will be available at https://github.com/Tear-kitty/MoEdit.

Problem

• Addresses inconsistent quantity perception in multi-object image editing.
• Proposes MoEdit for high-quality editing without auxiliary tools.
• Ensures object distinction and quantity consistency in complex edits.

Innovation

• Feature Compensation module minimizes object attribute interlacing
• Quantity Attention module ensures consistent quantity perception
• Leverages Stable Diffusion for high-quality multi-object editing