Video and Closed Caption



Key technologies and features:

  • Video Summarization: The video above is summarized version of Build 2016 Keynote. The summarization was made by using Azure Media Video Thumbnails media processor (MP). Original video source can be found at Channel9
  • Face detection and Identification: Speakers' names are displayed on the video closed caption above as the speakers acutally appear, which is achieved basically by the following steps - (1) registering speakers' faces in Cognitive Services' Face API beforehandin, (2) tuning video frames into thumbnails by using Azure Media Services' Media Encorder Standard, and (3) detecting and identifying speakers' faces by using Cognitive Servcies Face API
  • Jumping to a scene: You can jump to a scene where a speaker in the right column initially appears by clicking the name of speaker's face
Source Code: https://github.com/AzureMediaCognitiveDemos/VideoFramesFaceRecognition-Python