Improving advertising with eyetracking and deep learning

Test impact of video content instantly. Gain knowledge about user’s visual attention shifts. Identify visual attention errors. See Insight is a first of it’s kind visual attention prediction model for videos. It is based on deep learning and can evaluate video content in minutes. Model is trained with real eye tracking data therefore can predict visual attention of various users.

How it works

1. Select the video that you want to test

2. Get result

Application of the model

  • Optimize video advertisements - reach optimal visual attention
  • Implement eye tracking studies within minutes
  • Cut production process in half with instant insight’s to visual attention distribution
  • Easy improvement and adaptation of education material


  • Accountable results within minutes – model let’s you simulate thousands of users instantly. Validate and analyze your video narrative Fast and easy approach – it doesn’t require any participants or data collection

Where we are

  • We already have first versiono four algorithm and we’re now preparing for the next stage – adaptation to European markets.