I developed an interactive web-based Doodle Storyteller that combines
real-time machine learning with generative AI to create a fun, creative
experience for users. It allows users to draw up to three doodles, which are
recognized using DoodleNet (a pre-trained CNN). These recognized objects are
then sent as inputs to the Gemini API, which generates a unique and
imaginative short story based on the combination of doodled objects.