HYPERDREAM
2023
Hypnagogic landscapes generated from EEG signals
role
Concept, experiment design, EEG data collection and analysis, text-to-mesh generation
authors
Taisija Demchenko, Ninon Lizé Masclef
exhibitions
Creativity and Cognition 2023 virtual gallery
conference
Ninon Lizé Masclef and Taisija Demchenko. (2023). Hyperdream : Exploring Hypnagogia Organicity. In Creativity and Cognition (C&C '23). ACM, New York, NY, USA. DOI:10.1145/3591196.3593506
keywords
3D graphics, brain-computer interface, machine learning, neurosciences, immersive art, research
For the ACM Creativity & Cognition 2023 exhibition, Taja Demchenko and I created hypnagogic landscapes from EEG spectrograms. With the aid of a Dream Machine, the individual enters in a dream-like state. We generate 3D landscapes inspired from Frank Lloyd Wright's organic architectures by leveraging diffusion models (img2img and img2mesh).


Initial 2D image idea generation


Presentation video of hypngogic landscapes


Initial 3D generation from EEG spectrogram, using Stable Diffusion and ZoeDepth.


Hypnagogic landscapes (3D)