Courseiva

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

A retail store wants to use Azure Computer Vision to count the number of people entering a store from live video feeds. Which prebuilt Azure Computer Vision capability should they use?

⚠ Common exam trap

Test-takers frequently confuse 'Object Detection' with 'People Detection,' assuming the prebuilt object detection model can reliably count people, but Microsoft specifically offers People Detection as a separate, optimized API for this exact use case.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

People Detection

People Detection is a specialized prebuilt capability within Azure Computer Vision that is designed specifically to detect and count people in images or video frames. Unlike generic object detection, it is optimized for identifying human figures regardless of pose or occlusion, making it ideal for counting store entries from live video feeds.

Answer analysis

Option-by-option breakdown

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

  • Image classification

    Why it's wrong here

    Image classification assigns a single semantic label or a set of class scores to the entire image, such as 'store' or 'crowded scene', without localizing any objects. It cannot produce a per-person count because it never identifies individual instances, boundaries, or positions. Even if the classifier predicts the presence of people, that prediction is a whole-image attribute and carries no information about how many humans are actually present.

    When this WOULD be correct

    A question asking which capability to use for categorizing a store image as 'busy' or 'empty' based on overall scene content would make image classification correct.

  • Object detection

    Why it's wrong here

    Object detection identifies and locates multiple objects within an image using bounding boxes, but it does not track movement across frames or count unique individuals entering a doorway over time. This fails the scenario because the requirement is for a continuous count from live video, not a per-frame detection of static objects. It is tempting because object detection can locate people in a single frame, making it a natural choice for a one-off headcount in a still image, where it would be correct.

    When this WOULD be correct

    A question asking: 'Which Computer Vision capability should be used to identify and locate multiple product types on a shelf in a retail store?' would make object detection correct, as it can detect various object classes with bounding boxes.

  • People Detection

    Why this is correct

    People Detection is the prebuilt Azure AI Vision capability specifically trained to localize human bodies in images and video frames; it returns bounding boxes and a confidence score for every person it finds, so a count is simply the number of returned detections. Because the model is tuned for human figures, it handles partially obscured people and crowded retail scenes better than a general-purpose object detector, and it can be fed sequential frames to support visitor-count analytics. This purpose-built design is exactly why it is the correct choice for counting people.

  • Optical Character Recognition (OCR)

    Why it's wrong here

    Optical Character Recognition (OCR) is engineered exclusively to detect and extract text from images, returning machine-readable strings and their bounding polygons. It contains no visual model for human bodies or other non-text objects, so a crowd of people would simply produce no OCR output. Using OCR to count people is functionally impossible because the pipeline ignores everything that is not a character or word.

    When this WOULD be correct

    A question asking which Azure Computer Vision capability should be used to read text from store signage or receipts from live video feeds would make OCR the correct answer.

Option-by-option analysis

Why each answer is right or wrong

Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The AI-900 exam frequently reuses these exact scenarios with slightly different constraints.

People DetectionCorrect answer

Why this is correct

People Detection is the prebuilt Azure AI Vision capability specifically trained to localize human bodies in images and video frames; it returns bounding boxes and a confidence score for every person it finds, so a count is simply the number of returned detections. Because the model is tuned for human figures, it handles partially obscured people and crowded retail scenes better than a general-purpose object detector, and it can be fed sequential frames to support visitor-count analytics. This purpose-built design is exactly why it is the correct choice for counting people.

Image classificationWrong answer — click to see why

Why this is wrong here

Image classification assigns a single label to an entire image, but the task requires detecting and counting multiple people in a video feed, which is a spatial detection problem.

★ When this WOULD be the correct answer

A question asking which capability to use for categorizing a store image as 'busy' or 'empty' based on overall scene content would make image classification correct.

Why candidates choose this

Candidates may confuse image classification with object detection, thinking that classifying an image as 'person' is sufficient for counting, but classification does not provide locations or counts.

Object detectionWrong answer — click to see why

Why this is wrong here

Object detection identifies and locates objects in an image but does not specifically count people; it would require additional logic to filter for 'person' class and track across frames, whereas People Detection is optimized for counting people in crowds.

★ When this WOULD be the correct answer

A question asking: 'Which Computer Vision capability should be used to identify and locate multiple product types on a shelf in a retail store?' would make object detection correct, as it can detect various object classes with bounding boxes.

Why candidates choose this

Candidates may think object detection can count people by detecting 'person' objects, overlooking that Azure's People Detection is a specialized prebuilt model for accurate people counting in crowded scenes.

Optical Character Recognition (OCR)Wrong answer — click to see why

Why this is wrong here

OCR is designed to extract text from images, not to detect or count people. The question requires counting people entering a store, which is a people detection task, not text recognition.

★ When this WOULD be the correct answer

A question asking which Azure Computer Vision capability should be used to read text from store signage or receipts from live video feeds would make OCR the correct answer.

Why candidates choose this

Candidates may confuse OCR with general image analysis capabilities, thinking it can 'read' or identify objects in a scene, or they may not be aware of the dedicated People Detection feature.

Analysis generated from the official AI-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”

About these practice questions

This AI-900 question is part of Courseiva's 985-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

JA

Written by Johnson Ajibi, MSc IT Security

Senior Network & Security Engineer · founder of Courseiva

This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.