Spring 2021 Cognitive Computing Class Project
Need to write a haiku about something you just saw but not sure how to go about it?
My team and I worked on a project this semester that identifies the subject of an input image using a pretrained computer vision model (VGG16), generates a set of haikus using a transformer (GPT-2) and selects the haiku most associated with the image using OpenAI's Contrastive Language–Image Pre-training (CLIP).
Really enjoyed working on it with Pete Davis, Ram Kapistalam, India Lindsay, Alex McGraw and Matthew Streichler!
The link to an article describing our work: https://medium.com/geekculture/generating-haikus-from-images-c0d35c2470ce