To make something silly but fun, to use machine learning in a less mysterious way, to find interesting technologies and values

What it does

It takes in images and returns a haiku made with key terms appearing in the image

How we built it

Get tags from images through Clarifai API, count the number of syllables, fill up the 5-7-5 format with magic (by magic I mean a lot of work), read it aloud through Nuance API, develop an awesome user interface

Challenges we ran into

Debugging javascript, debugging angular.js and navigating through different APIs

Accomplishments that we're proud of

It delivers haiku in a brand new way that is extremely fun and entertaining

What we learned

More web dev, more javascript, using different APIs

What's next for Image to Haiku

Implement the feedback function for Clarifai API, so in the same time users can help improve the quality of service

Share this project: