Visually grounded emotion regulation via diffusion models and user-driven reappraisal

Edoardo Pinzuti^1,2Oliver Tüscher^1,2,3,4André Ferreira Castro^5*

• ¹Leibniz Institute for Resilience Research, Mainz, Germany
• ²Department of Psychiatry, Psychotherapy and Psychosomatic Medicine, University Medical Center Halle, Halle (Saale), Germany
• ³German Center for Mental Health (DZPG), Site Halle-Jena-Magdeburg, Halle (Saale), Germany
• ⁴Department of Psychiatry and Psychotherapy, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany
• ⁵School of Life Sciences, Technical University of Munich, Freising, Germany

Cognitive reappraisal is a key strategy in emotion regulation, involving reinterpretation of emotionally charged stimuli to alter affective responses. Despite its central role in clinical and cognitive science, real-world reappraisal interventions remain cognitively demanding, abstract, and primarily verbal in nature. This reliance on higher-order cognitive and linguistic processes can be especially impaired in individuals with trauma, depression, or dissociative symptoms, limiting the effectiveness of standard approaches. Here, we propose a novel, visually based augmentation of cognitive reappraisal by integrating large-scale text-to-image diffusion models into the emotional regulation process. Specifically, we introduce a system wherein users reinterpret emotionally negative im-ages via spoken reappraisals, which are then transformed into supportive, emotionally congruent visualizations using stable diffusion models with a fine-tuned IP-adapter module. This generative transformation visually instantiates users' cognitive reappraisals while maintaining structural similarity to the original stimuli, thus externalizing and reinforcing regulatory intent. To evaluate this approach, we conducted a within-subjects experiment (N = 20) using a modified cognitive emotion regulation (CER) task. Participants reappraised or described aversive images from the international affective picture system (IAPS), with or without AI-generated visual feedback. Re-sults indicate that AI-assisted reappraisal significantly reduced negative affect relative to both non-AI reappraisal and control conditions. Further analyses show that sentiment alignment between participant reappraisals and generated images correlates with affective relief, suggesting that multimodal coherence enhances regulatory efficacy. Our findings highlight the feasibility of using generative visual support for cognitive reappraisal. This work opens a new interdisciplinary direction at the intersection of generative AI, affective computing, and therapeutic technology design.