Advancing the generation and integration of traditional motifs through AI-based techniques

This study introduces an approach to blending abstract and representational art forms, focusing on two traditional Thai motifs: Ban Chiang and Sangkhalok. Using generative artificial intelligence (AI) techniques such as image fusion, text-guided image generation, and Low-Rank Adaptation (LoRA) fine-tuning, the research explores how modern technology can both preserve and enhance the cultural heritage of these motifs while incorporating contemporary design elements. Image fusion effectively generated Ban Chiang motifs with minimal computational resources, while LoRA fine-tuning significantly enhanced Sangkhalok motifs by introducing modern features. The fusion of Ban Chiang’s abstract qualities with Sangkhalok’s representational style benefited from LoRA fine-tuning, leading to improved motif integration and creativity, though challenges such as difficulty in controlling color accuracy were noted. This research serves as a foundational exploration of how AI can be a powerful resource for future cultural preservation and artistic innovation. While not a final design tool for artisans, it highlights the potential of AI-based generative techniques for preserving and innovating traditional motifs, emphasizing the need for further refinement in achieving precise blending of artistic forms.