ThoughtDiffusion: An Interactive Installation for Exploring Neuro-Art from EEG Data with Stable Diffusion Models
Year: 2025
Type: conference-paper
Abstract: ThoughtDiffusion is an interactive installation that combines cognitive processes with the generative capabilities of AI to generate artistic images in real time.The installation system uses Stable Diffusion models and state-of-the-art neural decoding techniques that allow the mapping of brainwave patterns into coherent visual representations.The installation is based on a non-invasive commodity EEG headset that records users' brain signals fed into a stable diffusion model to output images corresponding to the intended mental state but unique to the participants.This installation uses a Kinect V2 sensor to capture users' body movements, which advance multimodal interaction significantly relating to relaxation, the state of being calm, and the state of attention.
Author: R.P.C. Janaka Rajapakse
Institution: Tainan National University of the Arts
Language: English
Topic: Art, Technology, and Culture
Subfield: Visual Arts and Performing Arts
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W4409913540