Опубликована: Сен. 10, 2024

Mental health issues are prevalent and challenging both on a personal societal level. Art therapy has been recognized method of treatment, offering patients an expressive non-verbal way to process emotions. With advancements in digital technologies, integrating artificial intelligence (AI) into therapeutic practices the potential enhance accessibility effectiveness. However, application generative AI within art remains underexplored, necessitating further investigation its technical ethical challenges. This study aims develop showcase novel design that integrates therapy, specifically focusing use text-to-image models. The goal is provide supportive tool enhances patient expression creative customization, while preserving role therapist guiding process. We define simplified session demonstrate how it can be augmented with AI. Our implementation leverages edge detection, sketch-to-image models, inpainting techniques enable refine their artwork through text prompts. qualitatively evaluate system using three illustrative examples: sketch, painted image, photograph sculpture. successfully generated, refined adapted versions exemplary artworks for all cases, demonstrating ability maintain essence original input. integration offers significant enhancing possibilities patients. While this proof-of-concept highlights feasibility, future research required assess efficacy, user acceptance, address concerns, such as bias AI-generated content. approach foundation combining traditional practices, supporting development extended applications clinical settings. implementations publicly available at https://github.com/BFH-AMI/sds24.

Язык: Английский