How to Verify Only One Person Appears in a Live Camera Stream?
You are building an internet-based training solution. The solution requires that a user’s camera and microphone remain enabled. You need to monitor a video stream of the user and verify that the user is alone and is not collaborating with another user. The solution must minimize development effort. What should you include in the solution?
Community Insight
The test is recognizing a prebuilt real-time video analytics capability (people detection/counting) versus custom-trained object detection; the trap is assuming you must train an object detector with Custom Vision to detect a person.
This question asks which Azure AI service can monitor a live camera feed and confirm that only the user is present, with minimum development effort. Spatial Analysis in Azure Vision in Foundry Tools is the answer because it is a prebuilt, real-time video capability that detects and counts people without any custom model training.
Picking object detection in Azure AI Custom Vision because it sounds like 'detect a person in an image' — it works, but it demands labeled data, training, evaluation and a custom deployment, which directly violates the minimize-development-effort requirement.
Community Discussion (1 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
Spatial Analysis in Azure Vision in Foundry Tools is a prebuilt capability purpose-built for real-time streaming video, consuming an RTSP feed from the user's camera and emitting person-detection and people-counting events. Because person detection and zone-based occupancy are provided out of the box, you can assert that exactly one person occupies the frame and raise an alert when a second person enters — with configuration rather than model development. That is precisely what the requirement asks for: continuously monitor the stream, verify the user is alone, and minimize development effort. No labeled dataset, training run or custom container needs to be produced.Why the Other Options Are Wrong
Content moderation in Azure AI Video Indexer (A) is designed to flag offensive or inappropriate visual/audio content, and it produces moderation insights rather than live occupancy signals, so it cannot prove the user is alone. Speech-to-text in Azure Speech in Foundry Tools (B) transcribes spoken audio; it can tell you someone is talking but never whether a second body is in the camera's field of view. Object detection in Azure AI Custom Vision (C) can technically detect people, but it requires collecting and labeling images, training and iterating a model, and deploying it yourself — the most effort-heavy path in the list, and one that degrades when people overlap or lighting changes. Only Spatial Analysis gives ready-made, real-time person counting.Community Comment Notes
Srikantha argued for the same option, noting that Spatial Analysis is "a pre-built capability in Azure AI Vision specifically designed to analyze real-time streaming video" and that it can continuously count the number of people in the camera's view to verify only the user is present. That reasoning matches the exam intent: prebuilt real-time video analytics beats a custom-trained detector whenever the requirement explicitly says to minimize development effort. The practical takeaway from the discussion is to pair Spatial Analysis with a zone/occupancy rule and an alert, rather than building your own detection pipeline.Official Reference
Exam Strategy
Whenever a stem says "minimize development effort" and the scenario involves live video, scan the options for a prebuilt Azure AI Vision capability (Spatial Analysis) before considering anything that requires training a model. Reserve Custom Vision for cases where the object or condition is genuinely custom and no prebuilt detector exists.
Frequently Asked Questions
Why is Custom Vision object detection a poor fit for verifying the user is alone?
It requires labeled images, training, evaluation and a custom deployment, which adds effort the question explicitly wants to avoid, and accuracy drops with overlapping people.
Does Spatial Analysis need the camera stream in a specific format?
It ingests a real-time RTSP video feed from an IP camera, so the user's camera stream must be exposed as RTSP to the Spatial Analysis container or service.
Related Analysis
Practice All AI-103 Questions
Access 88 questions with complete answers and detailed explanations.
View Full AI-103 Practice Test →