When we joined DevHacks, we saw that Clozyt, a sponsor, was focusing on clothing apps. But since only teams of two could compete with the official Clozyt challenge, we decided to take the idea in a new direction that worked for our group of four.
Instead of building a shopping platform, we wanted to explore computer vision in fashion. Our project focuses on using AI to recognize clothing in images taken by the user and then recommending related items. It’s less about being a store, and more about showing how vision systems can power the future of fashion discovery.
We were inspired by the question: “What if your camera could understand your outfit?” From there, we built a camera tool, integrated an AI model, and developed a simple recommendation flow.
Along the way, we learned about AI models like YOLO and MobileNet, struggled with connecting them to the browser, and found creative workarounds to make a functional demo. We also had to rethink how to present recommendations without a large product database, so we tried to use APIs and mock data to simulate the experience.
Our biggest challenges were integrating vision AI in the browser and keeping the system lightweight enough for a hackathon setting.
Log in or sign up for Devpost to join the conversation.