StyleSense
An interactive smart mirror that gives personalized style advice using computer vision and AI.
Challenge
Many retail storefronts struggle to attract passers-by. Public spaces are sites for social exchange and cultural expression, yet most store experiences remain static and disengaging. We explored how interactive and playful technology could create more engaging customer experiences while also serving the store's need for constant renewal.
Challenge
Many retail storefronts struggle to attract passers-by. Public spaces are sites for social exchange and cultural expression, yet most store experiences remain static and disengaging. We explored how interactive and playful technology could create more engaging customer experiences while also serving the store's need for constant renewal.
Approach
StyleSense is an interactive mirror that analyzes what you're wearing and delivers personalized style advice through positive, engaging feedback. The backend uses Python with Google Gemini for real-time outfit analysis and recommendation generation, paired with a simple HTML interface housed in a physical mirror frame. The system ties its suggestions to actual products available in the store, shifting the interaction from isolated purchase moments toward long term brand loyalty.
Approach
StyleSense is an interactive mirror that analyzes what you're wearing and delivers personalized style advice through positive, engaging feedback. The backend uses Python with Google Gemini for real-time outfit analysis and recommendation generation, paired with a simple HTML interface housed in a physical mirror frame. The system ties its suggestions to actual products available in the store, shifting the interaction from isolated purchase moments toward long term brand loyalty.
Process
The project followed a Double Diamond process with observations, literature studies, and thematic analysis shaping the problem space. Prototyping moved through storyboards, user flows, and scenarios before reaching a high fidelity prototype. Usability testing with visual merchandisers and end users led to a final integration where StyleSense recommends catalog products tailored to the user's outfit. Co-created with Fabian Portada Vinberg, Samuel Stone, and Yousef Al Tall. I was responsible for the high fidelity prototype.
Process
The project followed a Double Diamond process with observations, literature studies, and thematic analysis shaping the problem space. Prototyping moved through storyboards, user flows, and scenarios before reaching a high fidelity prototype. Usability testing with visual merchandisers and end users led to a final integration where StyleSense recommends catalog products tailored to the user's outfit. Co-created with Fabian Portada Vinberg, Samuel Stone, and Yousef Al Tall. I was responsible for the high fidelity prototype.