Personalized adaptive generation of peking opera facial makeup using generative artificial intelligence

Area of Science:

• Artificial Intelligence
• Computer Vision
• Cultural Heritage Preservation

Background:

• Peking Opera facial makeup generation is challenging for AI.
• Existing methods lack high-fidelity detail and sufficient training data.

Purpose of the Study:

• To develop an AI model for generating detailed Peking Opera facial makeup.
• To address limitations in current stylized image generation techniques.

Main Methods:

• Enhanced Stable Diffusion framework with region-specific noise scaling.
• Attention-augmented U-Net for improved detail synthesis.
• LoRA fine-tuning for efficient, high-quality generation.

Main Results:

• Achieved superior performance over State-Of-The-Art (SOTA) models.
• Obtained excellent FID (16.34), KID (9.44), and SSIM (0.4912) scores.
• Demonstrated effective preservation of cultural authenticity and accelerated production.

Conclusions:

• The proposed AI framework successfully generates authentic Peking Opera facial makeup with high fidelity.
• The model enhances detail synthesis and generation efficiency.
• This approach bridges AI capabilities with cultural heritage preservation needs.