Ecologies of Vision is a of digital artwork that uncovers patterns and behavioral choreographies of space and objects that are embedded in our visual systems. The project uses custom AI workflows and latent diffusion models to explore dynamics of perception and motion in relation to human and machine intelligence.

The methodology leverages latent space conditioning for temporal and structural coherence across generated frames, using a custom-designed model tailored for specific behavioral patterns. Attention mechanisms modulate feature extraction, while shared noise patterns facilitate consistent visual outputs that align with predefined spatial-temporal rhythms. By sequencing rhythms observed in natural systems into synthetic representations, the work constructs speculative ecologies that challenge traditional distinctions between natural and artificial systems.

Generative video, 4K, 2:30min 

Exhibited: 8-16 June 2024, Eskiden Piksele at Samsun Cok Amacli Salon, Samsun, Turkey
Back to Top