Kamo IO, a YouTuber, has developed an AI-powered mirror to critique outfits before the wearer leaves the house. Recognizing his own fashion challenges, Kamo IO sought a technological solution to improve his style. The AI mirror assesses outfits and provides feedback, helping him avoid fashion mistakes.

The AI mirror offers critical feedback. In one instance, it suggested an outfit was suitable for staying inside, followed by constructive suggestions. Multimodal AI models, like the one used, process images and text for nuanced understanding.

The system uses an ESP32-CAM to capture an image, which is sent to a laptop for processing. The laptop transmits the image to a multimodal model, like one from ChatGPT, with a prompt to evaluate fashion choices critically and humorously.

To train the AI, Kamo IO used Pinterest, curating images of styles he liked. Providing diverse data is crucial to avoid biases and ensure accurate results. The AI’s response is converted into audible feedback using text-to-speech software.

For those not wanting to build their own mirror, Kamo IO has a web-based app with similar functionality. Alternatively, he suggests sticking with shorts and flip-flops.

An ESP32-CAM is a low-cost microcontroller board with a built-in camera for IoT projects. A multimodal AI model processes different data types simultaneously. The accuracy of the AI advice depends on the training data provided; more diverse data improves the AI’s ability to provide useful advice.