Skip to content

Implement computer vision solutions Questions

Practice questions for Implement computer vision solutions topic in Microsoft Certified: Azure AI Apps and Agents Developer Associate. 26 questions covering this domain.

26 questions9 easy11 medium6 hard
Q1
easy

A developer wants a chat model that can analyze uploaded images and answer follow-up questions about them. What should be deployed?

Q2
medium

An app sends a harmful prompt to an Azure OpenAI image generation model. What behavior should the developer expect?

Q3
easy

When using the Image Edit API, which input defines the exact region the model should modify?

Q4
medium

A mobile app needs quick image understanding and can trade away fine detail to reduce token use. Which image detail setting should be sent to the visi...

Q5
hard

An insurer uploads a dense engineering photo with small warning labels and asks a vision-enabled model to extract fine visual details. Which image det...

Q6
medium

A designer needs a generated product sticker image with no background so it can be placed over other artwork. Which configuration is required?

Q7
hard

A developer tries to send 12 images in one vision chat request. What should they do to stay within the documented API limits?

Q8
easy

You need to verify that a person presenting to a camera is physically present and not a spoof (photo or replay). Which Azure AI service feature should...

Q9
medium

You need to extract structured insights from long-form video files, including spoken content, on-screen text, and scene segmentation. Which Azure serv...

Q10
easy

Which Azure AI Vision Image Analysis feature returns short natural-language captions describing the contents of an image?

Q11
hard

Your vision-enabled chat app needs to analyze a 12 MP photo accurately, including small text in the background. Which input strategy is most appropria...

Q12
medium

Your application generates marketing images with Azure OpenAI's image generation API and needs square 1024x1024 outputs at the highest fidelity. Which...

Q13
medium

An image is too large to upload directly and is hosted in private blob storage. You want a vision-capable model to analyze it without making the blob ...

Q14
medium

A developer generates product catalog images and needs each output to have a transparent background so that designers can composite them onto differen...

Q15
hard

Your application must enforce a brand policy that prohibits competitor logos in AI-generated images, and must also apply a watermark to all outputs fo...

Q16
medium

A medical imaging application must accurately detect fine anatomical labels in high-resolution scans sent to a vision-enabled chat model. Which image ...

Q17
hard

A security team discovers that users are embedding malicious instructions inside images (e.g., text overlaid on a photo) to manipulate a vision-enable...

Q18
easy

Which Azure OpenAI model family supports analyzing images as part of a chat conversation, allowing the model to answer questions grounded in visual co...

Q19
medium

An application generates images and must detect whether any output contains unsafe or prohibited visual content before serving it to users. Which AI-1...

Q20
easy

When using the Azure OpenAI Image Edit API to modify only a specific region of an image, which input determines the region to modify?

Sign in to see all 26 questions

Create a free account to browse all questions — completely free during our launch phase.