Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision
Summary
This paper introduces a hierarchical taxonomy and dense supervision training strategy for diffusion-based image editing, leveraging a large dataset of 12 million editing pairs and a granular evaluation benchmark to enhance model performance.
View Cached Full Text
Cached at: 08/25/26, 04:33 AM
Paper page - Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision
Source: https://huggingface.co/papers/2608.16812
Abstract
A hierarchical taxonomy and dense supervision strategy improve diffusion-based image editing through fine-grained concepts, large-scale paired data, and granular evaluation.
Existing image editing frameworks predominantly follow the training paradigm oftext-to-image diffusion models. However, extending this paradigm to image editing highlights two inherent discrepancies, specifically, the insufficient attention toedit concept granularityand the training inefficiency caused bysparse supervision signals. To address these issues, we establish a comprehensivehierarchical taxonomyfeaturing over 1,000 fine-grained edit concepts and buildConceptEdit-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 adense supervision training strategythat synthesizes multiplenon-interfering conceptsinto 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 presentConceptEdit-Bench, a granular evaluation suite designed to diagnose model capabilities across a vast array of real-world scenarios.
View arXiv pageView PDFGitHub5Add to collection
Get this paper in your agent:
hf papers read 2608\.16812
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2608.16812 in a model README.md to link it from this page.
Datasets citing this paper2
#### inclusionAI/ConceptEdit-12M Updatedabout 3 hours ago • 178 • 2 #### inclusionAI/ConceptEdit-Bench Preview• Updatedabout 20 hours ago • 1
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2608.16812 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Delta-Adapter: Scalable Exemplar-Based Image Editing with Single-Pair Supervision
Delta-Adapter enables exemplar-based image editing using single-pair supervision by extracting semantic deltas from pre-trained vision encoders and injecting them via Perceiver-based adapters, improving accuracy and generalization.
Learning 3D Editing without Paired Supervision via Generative Prior Distillation
This paper introduces a feed-forward 3D editing framework that uses generative prior distillation from foundation models to avoid paired training data, improving instruction fidelity and cross-view consistency.
HP-Edit: A Human-Preference Post-Training Framework for Image Editing
HP-Edit introduces a post-training framework that aligns diffusion-based image editing models with human preferences via RLHF, using a new 50K real-world dataset and an automatic VLM-based evaluator.
Abra: Scaling Diffusion Image Training
This paper presents a systematic scaling law study for text-to-image diffusion models, showing they scale predictably but require significantly more data per parameter than language models for optimal training.
Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning
Uni-Edit proposes using intelligent image editing as a single general task to simultaneously improve unified multimodal models' understanding, generation, and editing capabilities, with an automated data synthesis pipeline creating complex editing instructions.