Emotion detector

Problem:

  • People with Autism, Aspbergers or various other communication difficulty might have trouble recognizing emotions when interacting with others, and this can cause a lot of anxiety for these people going out into public and needing to deal with stranger interactions.

Solution:

  • Machine learning can detect what kinds of emotions someone is feeling based on the way that they're speaking - a combination of both the words they're using as well the intonation and way in which they're speaking.
  • Using Google's Machine Learning suite, we're able to record audio input, translate it to dozens of different languages, and then use natural language processing to determine what kind of emotion is being conveyed in the speach.

  • We've made an application that people can use to record conversations, so that they can get feedback about the emotions of others realtime and be able to better understand a conversation or daily interaction.

Share this project:

Updates