The inspiration from this project came from an introduction keynote at an MIT hackathon, talking about improving the human condition with technology.
What it does
The Image 2 Haiku website allows the user to upload an image to the website and submit it for analysis. The website then submits the image to multiple APIs including the Google Cloud Platform and Clarifai to generate a list of words describing it, analyze the weather in the location the image was taken, and analyze the mood and emotion of the image. These three factors are then brought together and utilized to generate a haiku that reflects the subject, weather, and mood of the image. As they say, "a picture is worth a thousand words."
How we built it
We built it using more than four different online APIs as well as the Python Flask library and PIL image analysis library. Each person took responsibility for one distinct API and became an expert in it, implementing it themselves and allowing us to come together at the end to pull them together into one unified program.
Challenges we ran into
We ran into some challenges implementing various APIs because of issues installing and utilizing them. However, working together as a team, we managed to overcome those difficulties and implement them.
Accomplishments that we're proud of
We were able to complete over 4 of the challenges offered with our application, including various weather, image recognition, and social good challenges. We also did not know each other before this hackathon, so we are proud of our ability to work as a team and develop an advanced product together.
What we learned
We learned how to use iterative design as well as work together as a team to produce a major product. We learned valuable teamwork skills and had frequent meeting to discuss both our progress and what still needed to be done, as well as determine whether anyone needed help. Overall, very valuable teamwork and development skills were gained from this.
What's next for Image 2 Haiku
If we have the chance to work more on this project, we had thrown around different ideas including integrating the Snapchat API to allow this to be directly used from Snapchat, implementing a refresh button to generate a fresh new poem without having to refresh, and improving the user frontend even more.