Flux-Sculptor: Text-Driven Rich-Attribute Portrait Editing through Decomposed Spatial Flow Control

2025-07-05

AI Summary

Text-driven portrait editing faces the dual challenge of imprecise spatial localization and content distortion, hindering simultaneous achievement of high reconstruction fidelity and flexible semantic editing. To address this, we propose a decoupled spatial flow control framework. First, we design a Prompt-Aligned Spatial Locator to achieve semantically grounded, high-precision region localization. Second, we introduce a Structure-to-Detail Edit Control strategy that hierarchically guides the diffusion denoising process, jointly optimizing structural consistency and photorealistic detail generation. Built upon the Flux architecture, our method integrates mask-guided latent representations with attention-value modulation to enhance spatial flow controllability. Extensive experiments demonstrate that our approach significantly outperforms state-of-the-art methods in fine-grained multi-attribute editing (e.g., hair color, makeup, age) while preserving facial identity and geometric fidelity. It achieves both high-fidelity reconstruction and robust editing performance, making it suitable for practical portrait editing applications.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Prompt Engineering / Prompting

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: Agentic search

Abstract

Text-driven portrait editing holds significant potential for various applications but also presents considerable challenges. An ideal text-driven portrait editing approach should achieve precise localization and appropriate content modification, yet existing methods struggle to balance reconstruction fidelity and editing flexibility. To address this issue, we propose Flux-Sculptor, a flux-based framework designed for precise text-driven portrait editing. Our framework introduces a Prompt-Aligned Spatial Locator (PASL) to accurately identify relevant editing regions and a Structure-to-Detail Edit Control (S2D-EC) strategy to spatially guide the denoising process through sequential mask-guided fusion of latent representations and attention values. Extensive experiments demonstrate that Flux-Sculptor surpasses existing methods in rich-attribute editing and facial information preservation, making it a strong candidate for practical portrait editing applications. Project page is available at https://flux-sculptor.github.io/.

Problem

• Precise text-driven portrait editing localization
• Balancing reconstruction fidelity and editing flexibility
• Rich-attribute editing with facial information preservation

Innovation

• Flux-based framework for precise portrait editing
• Prompt-Aligned Spatial Locator for accurate region identification
• Structure-to-Detail Edit Control for guided denoising