Inspiration
Often I share giffs with friends and family to express emotions that could not otherwise be expressed in words. The motivation of the project was to use image recognition to classify emotion of 'selfies' and to recommend to you giffs to share (from giphy) to others to let them know how you are feeling.
What it does
Take a picture with your camera, or upload one from your camera library. Instead of classifying emotion, the project went a slightly different direction. Given any image that you share, the application recommends to you giffs that are proportional to what you see in front of you.
How I built it
Using the Microsoft Computer Vision API and the giphy search api. The client communication with these services is all done in html/css/js. Using firebase for hosting.
Challenges I ran into
Convincing my teammates to help (they just wanted to do homework!)
Accomplishments that I'm proud of
It is not the most beautiful thing in the world but I am impressed that it works!!
What I learned
I learned that you never want to screw around with 'octet-streams' when needing to pass an image in a post request.
What's next for mo0w
Refining the recommendation service. Creating several more features (including catering more to the original idea). For the future, enable those who may have a difficult time understanding other peoples emotions (based on non-verbal cues i.e. some on the Autistic Spectrum) by allowing them to take pictures and see expressions in movement (through playful giffs) to aid in understanding what others may be feeling.
Log in or sign up for Devpost to join the conversation.