NoHumansRequired: Autonomous High-Quality Image Editing Triplet Mining

2025-07-18

AI Summary

High-quality image editing triplets (source image, instruction, edited image) for supervised training of generative models are scarce and difficult to construct automatically—constrained by pixel-level accuracy, region controllability, style consistency, physical plausibility, and visual fidelity. Method: We propose the first fully automated, modular, high-fidelity triplet mining pipeline. It introduces a task-tuned Gemini-based validator jointly with image inversion and compositional self-bootstrapping to enable end-to-end, unsupervised evaluation of instruction adherence and aesthetic quality—eliminating reliance on segmentation models or human annotation. Contribution/Results: We release NHR-Edit, a large-scale dataset comprising 358K triplets, and Bagel-NHR-Edit, a fine-tuned diffusion model. Extensive cross-dataset evaluation demonstrates state-of-the-art performance, surpassing all existing methods across multiple benchmarks.

Technology Category

Computer Vision: Diffusion Models for VisionNatural Language Processing: GenerationMachine Learning: Large Multimodal Models (LMMs)

Application Category

Economics, Online Markets and Human Computation: LLM based quality controls for crowd workSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingWeb Mining and Content Analysis: Large pretrained models with web data

Abstract

Recent advances in generative modeling enable image editing assistants that follow natural language instructions without additional user input. Their supervised training requires millions of triplets: original image, instruction, edited image. Yet mining pixel-accurate examples is hard. Each edit must affect only prompt-specified regions, preserve stylistic coherence, respect physical plausibility, and retain visual appeal. The lack of robust automated edit-quality metrics hinders reliable automation at scale. We present an automated, modular pipeline that mines high-fidelity triplets across domains, resolutions, instruction complexities, and styles. Built on public generative models and running without human intervention, our system uses a task-tuned Gemini validator to score instruction adherence and aesthetics directly, removing any need for segmentation or grounding models. Inversion and compositional bootstrapping enlarge the mined set by approximately 2.2x, enabling large-scale high-fidelity training data. By automating the most repetitive annotation steps, the approach allows a new scale of training without human labeling effort. To democratize research in this resource-intensive area, we release NHR-Edit: an open dataset of 358k high-quality triplets. In the largest cross-dataset evaluation, it surpasses all public alternatives. We also release Bagel-NHR-Edit, an open-source fine-tuned Bagel model, which achieves state-of-the-art metrics in our experiments.

Problem

• Automated mining of high-quality image editing triplets without human intervention
• Lack of robust metrics for evaluating edit quality at scale
• Need for large-scale training data without human labeling effort

Innovation

• Automated pipeline mines high-fidelity image triplets
• Uses Gemini validator for quality scoring without segmentation
• Inversion and bootstrapping expand dataset by 2.2x