Quickmoji: Most Innovative Hack - Hack@Brown Emotion Recognition to Speed up Texting

Emojis nowadays are our go-to means to inject some humanity into a conversation or quickly react to something, whether it is in text, chats, or instagram.

The problem is, it’s still not that human, especially if you’re stuck in a meeting and you can’t just send a “thumbs up” emoji to your boss that just gave you a to-do list for tomorrow. And it’s definitely not that quick, especially if you’re [again], stuck in a meeting and you can’t just send an emoji to a friend that just frantically texted “wya”.

To top that off, messenger apps give us emojis and text suggestions that often do not make sense and are not tailored to each situation.

When: January, 2020
Client: Hack@Brown, 2020
Role: Product Management
Tools: Google Cloud API, Figma

Introducing Quickmoji: where you can respond to messagers quicker with real-time emotion recognition.

Customized text replacement for each Emoji

Similar to text replacement, each emoji would correspond to a quick action, such as, texting “sounds good” or even sending your location.

"Ding"

“Ding”: hold down message for “quick reactions,” which is when your phone would perform an emotion recognition scan.

Face scan that gives you tailored quick reacts.

This would generate the top four emotions that match your face currently, which correspond to different quick actions like sending a "sounds good!" or an "On my way!" text.

Demo Day

Demo Day At the hackathon, we were able to show proof-of-concept with a front-end web demonstration. We created an archive of 20 emojis, with each emoji matched to values ranging from 0 to 4 on the scales of Joy, Sorrow, Anger, and Surprise, which are the scales on which the Google Cloud API rates emotion. We then built a nearest-neighbor classification algorithm with each emoji embedded in 4D space that matched to emotions the facial recognition API detected. For the front end, we visualized of program features and prototyped with Adobe XD.