Micro Fashion Network

A network map of street fashion—tracking how color combinations surface, cluster, and circulate in public space.

Micro Fashion Network maps the evolution of street-level fashion trends by tracing the dominant colors of clothing worn by passersby over time. Set in Harvard Square, Cambridge, the project focuses on two primary elements: color and time.

Using a fixed-position camera and custom software, the work captures people walking through a busy public space. The system extracts and stores the dominant clothing colors at regular intervals, building a growing archive of color samples. As time progresses, similar color combinations ared linked, forming a complex network of recurring hues.

The resulting network follows a preferential attachment model, producing a power-law distribution—a few dominant colors emerge as central hubs, while others remain peripheral. This structure offers a visual language for understanding how color combinations in clothing emerge, gain traction, and fade within everyday urban life.

The final presentation takes three complementary forms:

1. Captured human figures – snapshots of individuals recorded by the camera.
2. Abstract box compositions – visualizations of extracted color data.
3. A dynamic network graph – mapping how color combinations connect and cluster over time.

Micro Fashion Network is a generative system for observing the relational aesthetics of subtle shifts that shape fashion in the rhythms of daily life.