ReactTube

• 

  Reaction emotion capturing

• 

  Video analytics page

• 

  Main page

Inspiration

• Content creators would like to have a better insight of what kind of emotions their viewers experience when watching.
• Viewers would like to see what emotions other viewers had when watching certain videos/films.
• Viewers would also like to skip to certain parts of the video they are most interested in, and can do this based on emotional reaction to the part of the video.
• New video recommendations can be made to users based on user mood/emotions.
• It is interesting to compare emotion data with different kinds of users with different demographics.

What it does

• ReactTube displays a video to the user, while recording the user's reaction. The app uses AI to analyze the user's emotions during the course of the video, and uploads the data to the web site. The website allows users to log on and view videos, and the reactions associated with them. The play bar on each video is color-coordinated with different kind of emotions expressed by viewers, for each second of the video.

How we built it

• Microsoft Azure Emotions API
• .NET Framework
• Node.JS
• Chart.JS

What's next for ReactTube

• More insights
• More data
• Content creators could learn from the intel gathered from ReactTube data.
• Once enough data is gathered, it would be interesting to compare different demographics of people and not only supply the data to others, but analyze the data and other aspects to determine why certain demographics react and express emotions certain ways, while others wouldn't.

Built With

• .net
• ai
• azure
• chart.js
• cognitive-system
• computer-vision
• machine-learning
• node.js
• youtube