Microsoft Emoji8 uses machine learning to judge emoji imitation

Brittany A. Roston - Nov 16, 2018, 5:29 pm CDT
Microsoft Emoji8 uses machine learning to judge emoji imitation

Microsoft has unveiled a fun new app that utilizes its machine learning technology to judge facial expressions imitating emoji. The Emoji8 tool is a sample UWP application, according to Microsoft, that uses a device’s webcam to monitor users imitating randomly selected emoji expressions. Emoji8 is provided both through the Windows Store and on GitHub.

Emoji8 is described as a “fun sample UWP app” that utilizes Windows Machine Learning and presents it in a way that is easy for anyone to experience. Users are presented with a randomly chosen emoji that they must imitate with their own facial expression.

The app uses a webcam to monitor the user’s face and evaluate their expression relative to the emoji. A rating is given for each attempted imitation and users have the option of sharing their best results as a GIF via Twitter.

The app is open-sourced on GitHub for those who are interested in tinkering with the code; everyone else will need to grab it through the Windows Store using a PC running Windows. The app will record a video feed from the user’s computer and use FER + Emotion Recognition Model version 1.2 to evaluate the content.

This evaluation takes place locally on the computer, which should help mitigate any data security concerns. An Internet connection is not necessary to use the app, but it is obviously needed to directly tweet the GIF results. The app can be downloaded on Windows Store and GitHub now.

Must Read Bits & Bytes