Beyond Facial Consistency: Personalized Person Image Generation with Holistic Identity Preservation

2026-07-28

AI Summary

Existing methods for personalized portrait generation struggle to simultaneously preserve fine facial details and maintain global appearance consistency, often resulting in an imbalance of identity information across local and global scales. This work proposes a dual-branch generative framework that separately models global appearance and local facial features, complemented by a novel Dynamic Balance Scaling (DBS) strategy. DBS employs adaptive temporal gating and region-aware optimization to dynamically coordinate the contributions of both branches, effectively mitigating the dominance of the facial branch. Additionally, the authors introduce the Pexels-100 benchmark for evaluation. Experimental results demonstrate that the proposed approach achieves a superior trade-off between facial fidelity and identity consistency, significantly outperforming existing open-source baselines and establishing a foundational framework for controllable full-identity modeling.

Abstract

Personalized person image generation requires preserving subject identity across both local facial details and broader appearance cues. Existing methods typically emphasize only one level of identity information, leading to an inherent trade-off between facial fidelity and overall appearance consistency. To address this, we first propose a simple dual-branch baseline that unifies global appearance control and local facial control within a shared generation framework. This simple combination of different branches yields promising results, but suffers from instability in practice due to uncoordinated branch contributions. To this end, we propose Dynamic Balancing Scaling (DBS), a fine-tuning strategy for improving face and appearance identity coordination. DBS consists of two components: adaptive temporal gating, which dynamically modulates branch contributions along the denoising trajectory, and region-aware optimization, which improves the coordination of facial, appearance, and global supervision. Together, these designs alleviate persistent face-branch over-dominance and encourage more effective appearance-aware guidance. We also introduce Pexels-100, a benchmark for evaluating holistic identity consistency in personalized person generation. Experiments show that DBS achieves a better trade-off between facial fidelity and appearance consistency than existing open-source baselines, while providing a controllable basic framework for holistic identity modeling.

Problem

• personalized person image generation
• identity preservation
• facial fidelity
• appearance consistency
• holistic identity

Innovation

• Dynamic Balancing Scaling
• dual-branch generation
• holistic identity preservation
• region-aware optimization
• personalized image generation