Math + Code + Art Initiative · UCLA Mathematics · 2026
Cellular Impressions
Portraits Reimagined Through Algorithms
Tucker Nielson, Weimo Zhu, Jiayin Lu, Ying Jiang, Michael Andrews, Chenfanfu Jiang
Artist Statement
Cellular Impressions explores identity and the formation of visual impressions. Each work begins with a photographic portrait. The image is then transformed using neural style transfer, which introduces visual textures and color patterns drawn from other images. After that, adaptive Voronoi tessellation divides the image into geometric cells, fragmenting the portrait while still preserving its overall structure.
The resulting images appear both familiar and abstract. From a distance, viewers recognize the person in the portrait. Up close, the image dissolves into geometric regions and painterly textures. This layered transformation reflects how impressions of people are often formed through interpretation rather than direct perception.
By combining facial expression with contrasting visual styles, the series invites viewers to think about where a person's identity seems to reside — in the face itself, in the artistic interpretation, or in the viewer's perception.
Mathematics and code serve as creative media in this work. Through geometric partitioning and neural style transfer, computational processes translate abstract ideas about identity and perception into visual form.
Works