Every morning, we make unconscious choices about what to wear. These decisions — the colors, textures, and layers we reach for — carry emotional information we rarely examine. TAMED began with a personal need: to understand my own emotional landscape by observing the one behavior I perform every day without thinking. I wanted to build a machine that could read my outfit choices and reflect back the feelings I hadn't yet acknowledged.
This wasn't about fashion prediction or style recommendation. It was about using clothing as emotional data — a diary written in color and fabric rather than words. Could a system trained on my wardrobe choices learn to tame the unexisting consciousness, calming feelings before I even noticed them?
What if the colors you wear are a language your subconscious speaks fluently but your conscious mind can't read?
What if a machine could observe your patterns and offer healing before you realize you need it?
What if emotional self-awareness begins not with introspection, but with observation of everyday habits?
For months, I photographed my outfit every morning and kept a parallel diary of my emotional state. This dual dataset — visual and textual — became the training ground. I began to see patterns I'd never noticed: certain color combinations appeared consistently during anxious periods, while others emerged during calm. The act of observation itself became therapeutic, but I wanted to go further.
Using machine learning, I built a system that analyzed the color composition of each outfit and correlated it with diary entries. The machine learned associations between color palettes and emotional states — not universal truths, but my personal emotional vocabulary expressed through clothing. The most surprising discovery was how accurately it could predict my mood from colors alone.
The final output is a "tamed quote" — a personalized message generated from the intersection of what I wore and what I felt. These quotes don't diagnose or prescribe; they reflect. They hold up a mirror to an emotional state I may not have recognized, offering a moment of subconscious healing through acknowledgment. Exhibited at PARALLEL, the installation invited viewers to consider their own unconscious emotional expressions.
A journal requires conscious reflection — you have to know what you feel before you can write it down. TAMED works in the opposite direction: it observes behavior and surfaces emotion. The machine sees patterns in my choices that I'm too close to recognize. It acts as an emotional translator, converting the visual language of clothing into words that acknowledge what I'm going through.
The installation format was essential because it externalized the process. Seeing my emotional data displayed — outfit photographs alongside generated quotes — created a distance that made self-understanding possible. The machine wasn't replacing human intuition; it was extending it into territory where consciousness hasn't yet arrived.
The data confirmed what I intuited but couldn't prove: outfit choices carry statistically significant emotional information. Specific color combinations correlated strongly with diary-documented emotional states, suggesting we dress our feelings before we feel them.
The act of receiving a tamed quote — a reflection of emotions I hadn't consciously processed — created a sense of being understood. Several viewers at PARALLEL described a similar experience when seeing the installation, recognizing their own unconscious emotional habits.
TAMED demonstrated that machine learning can be a tool for emotional intelligence rather than emotional replacement. The machine didn't understand my feelings — it surfaced patterns that helped me understand them myself.
TAMED taught me that the most powerful design interventions don't ask users to change behavior — they make invisible behavior visible. This principle now guides my professional work: the best interfaces surface what users already know but haven't articulated.
As my thesis project, TAMED also represents my deepest exploration of the relationship between data and human experience. The quantified self movement often reduces people to numbers; this project attempted the opposite — using numbers to recover the full emotional complexity of a person.
We dress our feelings before we feel them — TAMED taught me that the most honest diary I've ever kept was my closet.