Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision

2026-08-17

AI Summary

This study addresses the challenges of insufficient concept granularity and training inefficiency caused by sparse supervision in image editing. To overcome these limitations, we construct a ten-million-scale fine-grained editing dataset with a hierarchical taxonomy and propose a dense supervision strategy. By synthesizing multiple non-interfering concepts into single image pairs to augment learning signals, this approach effectively mitigates data distribution collapse. Experimental results demonstrate that our method significantly outperforms state-of-the-art techniques, simultaneously improving training efficiency and editing performance. Consequently, this work establishes an efficient data construction paradigm and a dedicated evaluation benchmark for fine-grained image editing, offering a robust solution to current data-centric bottlenecks in the field.

Abstract

Existing image editing frameworks predominantly follow the training paradigm of text-to-image diffusion models. However, extending this paradigm to image editing highlights two inherent discrepancies, specifically, the insufficient attention to edit concept granularity and the training inefficiency caused by sparse supervision signals. To address these issues, we establish a comprehensive hierarchical taxonomy featuring over 1,000 fine-grained edit concepts and build ConceptEdit-12M, a massive dataset of 12 million high-quality editing pairs via an improved synthesis framework. This library-driven approach effectively rectifies the distribution collapse of generated data while ensuring high data fidelity. Furthermore, we propose a dense supervision training strategy that synthesizes multiple non-interfering concepts into single image pairs. By providing richer learning signals, this strategy significantly enhances both training efficiency and overall model performance. Training results validate our strategy, significantly outperforming prior works. Finally, we present ConceptEdit-Bench, a granular evaluation suite designed to diagnose model capabilities across a vast array of real-world scenarios.

Problem

• Image Editing
• Concept Granularity
• Sparse Supervision
• Training Efficiency

Innovation

• Dense Supervision
• Fine-grained Edit Concepts
• ConceptEdit-12M
• Hierarchical Taxonomy
• ConceptEdit-Bench