LCP
Sports

Artificial intelligence-based solution for win prediction by using real-time game footage.

Client Profile

The client is a Europe-based sports organizer, arranging various racing events daily.

Challenge

  1. Race organizers and the management team were unable to analyze the motion parameters of racers.
  2. Viewers of the game had no means of forecasting the race result.

Solution

  1. Solved the result forecasting issue using artificial intelligence.
  2. Used Kinesis video stream to fetch real-time game streams from a video camera.
  3. Real-time stream is then passed onto AWS Lambda service, for video properties segmentation.
  4. Various video properties are processed separately.
  5. AI and machine learning algorithms are applied on frames collected for predicting the result.
  6. A boundary box is created for each race participant. Motion parameters and win probability (%) are displayed above each boundary box.

Key Benefits

  1. Motion parameter data collected helps the sports technologist and coach to guide the player with performance improvements.
  2. The solution helped the viewers to predict the game outcome.

Used technologies

Develop your next idea with usGet in touch