SelfAge: Personalized Facial Age Transformation Using Self-reference Images

2025-02-19

AI Summary

Existing facial age transformation methods model only population-averaged aging patterns, failing to capture individual-specific appearance variations arising from unique life histories—thus limiting personalization fidelity. This paper introduces the first diffusion-based personalized age transformation framework: leveraging 3–5 cross-age self-reference images of the same subject, it fine-tunes pre-trained diffusion models (e.g., Stable Diffusion) via a self-reference adaptation mechanism and customized conditional prompts—integrating identity anchoring and semantic age guidance. The method achieves natural, biologically plausible age editing while preserving strong identity consistency. Quantitative evaluation shows significant improvements over state-of-the-art methods (LPIPS reduced by 12.6%, ID similarity increased by 9.3%); qualitative assessment further confirms superior visual realism and personalization. The framework enables lightweight, controllable, and high-fidelity personalized age transformation.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search
• Security and Privacy: Privacy-enhancing technologies

Abstract

Age transformation of facial images is a technique that edits age-related person's appearances while preserving the identity. Existing deep learning-based methods can reproduce natural age transformations; however, they only reproduce averaged transitions and fail to account for individual-specific appearances influenced by their life histories. In this paper, we propose the first diffusion model-based method for personalized age transformation. Our diffusion model takes a facial image and a target age as input and generates an age-edited face image as output. To reflect individual-specific features, we incorporate additional supervision using self-reference images, which are facial images of the same person at different ages. Specifically, we fine-tune a pretrained diffusion model for personalized adaptation using approximately 3 to 5 self-reference images. Additionally, we design an effective prompt to enhance the performance of age editing and identity preservation. Experiments demonstrate that our method achieves superior performance both quantitatively and qualitatively compared to existing methods. The code and the pretrained model are available at https://github.com/shiiiijp/SelfAge.

Problem

• Personalized facial age transformation
• Individual-specific appearance preservation
• Diffusion model with self-reference images

Innovation

• Diffusion model for age transformation
• Self-reference images for personalization
• Fine-tuning with 3-5 images