AI-900 Practice Question: Describe features of computer vision workloads on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Azure AI Custom Vision

  Why it's wrong here

  Azure AI Custom Vision is meant for creating custom image classifiers and object detectors from tagged example images, and it processes one still image at a time. It does not natively provide temporal or audio signals, so it cannot track objects as they move across video frames, identify known people in a meeting, or transcribe spoken words. Since the scenario demands video-level analysis with participant identification, object tracking, and transcription, Custom Vision is not a suitable answer.

• ✓

  Azure AI Video Indexer

  Why this is correct

  Azure AI Video Indexer is the correct choice because it ingests video and audio and automatically extracts actionable insights using multiple integrated AI models. It can identify and track people and faces over time, detect objects, recognize scenes, and generate a time-stamped transcript of spoken dialogue, all from a single video file. This end-to-end video analysis makes it ideal for reviewing meeting recordings, whereas the other options address only narrow image or face tasks.

• ✗

  Azure AI Face

  Why it's wrong here

  Azure AI Face provides face detection, verification, and recognition on individual images, and it can return attributes like age, emotion, and head pose. It lacks the video-level capabilities of Video Indexer such as correlating faces across a shot, tracking a person throughout the recording, detecting scenes, or producing transcripts from speech. Thus, using Azure AI Face alone would fail to meet the full requirement of analyzing meeting videos for participant identification, object tracking, and transcription.

• ✗

  Azure AI Vision OCR

  Why it's wrong here

  Azure AI Vision OCR is designed to extract printed and handwritten text from still images, not to perform temporal analysis of video. Although Video Indexer also leverages OCR to read on-screen text from video frames, the Vision OCR service alone cannot identify meeting participants, track moving objects across frames, or generate an audio transcript. Therefore, it is a component of video analysis, not a complete video-analytics solution for the described need.

Go deeper

Related to this question