Question

A media company needs to build a searchable index of thousands of recorded training videos. Users must be able to find moments where a specific product name is either spoken by the presenter or displayed as on-screen text in slides. Which Azure AI Content Understanding capability set should you configure for the video analysis pipeline to satisfy BOTH requirements?

Accepted Answer

Enable speech transcription and on-screen OCR extraction, then index both the transcript timestamps and detected text regions

Answer

Enable only speech transcription, since presenter narration will always mention any product shown on screen

Answer

Enable shot/scene segmentation and thumbnail generation, then let users browse keyframes visually

Answer

Enable face detection and celebrity recognition to tag speakers, then search by speaker identity