Kontinuous Kontext: Continuous Strength Control for Instruction-based Image Editing

2025-10-09

AI Summary

Existing instruction-based image editing methods lack fine-grained, continuous control over editing intensity, preventing users from flexibly adjusting the degree of modification—from “no change” to “full edit”—via natural language instructions. To address this, we propose the first unified framework for continuous intensity control, introducing a learnable scalar intensity parameter mapped into the model’s modulation space via a lightweight projection network—enabling multi-attribute editing (e.g., style, attribute, texture, background, and shape) without retraining for each attribute. Leveraging state-of-the-art generative models, we construct a high-quality quadruple dataset (image–edit–instruction–intensity) with an integrated filtering mechanism to ensure data reliability. Experiments demonstrate that our method achieves high-fidelity, progressively controllable edits across diverse tasks, significantly enhancing users’ precision in regulating editing intensity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Instruction-based image editing offers a powerful and intuitive way to manipulate images through natural language. Yet, relying solely on text instructions limits fine-grained control over the extent of edits. We introduce Kontinuous Kontext, an instruction-driven editing model that provides a new dimension of control over edit strength, enabling users to adjust edits gradually from no change to a fully realized result in a smooth and continuous manner. Kontinuous Kontext extends a state-of-the-art image editing model to accept an additional input, a scalar edit strength which is then paired with the edit instruction, enabling explicit control over the extent of the edit. To inject this scalar information, we train a lightweight projector network that maps the input scalar and the edit instruction to coefficients in the model's modulation space. For training our model, we synthesize a diverse dataset of image-edit-instruction-strength quadruplets using existing generative models, followed by a filtering stage to ensure quality and consistency. Kontinuous Kontext provides a unified approach for fine-grained control over edit strength for instruction driven editing from subtle to strong across diverse operations such as stylization, attribute, material, background, and shape changes, without requiring attribute-specific training.

Problem

• Enables fine-grained control over image edit strength through instructions
• Extends editing models to accept scalar inputs for gradual adjustments
• Provides unified strength control across diverse editing operations without retraining

Innovation

• Adds scalar strength control to text instructions
• Trains lightweight projector for modulation coefficients
• Synthesizes quadruplet dataset for strength training