Azure AI Content Safety Image Severity Output
Short answer
A The Azure AI Content Safety API returns a severity level of 0 for a non-offensive image when using the EightSeverityLevels output type.
Question
Key Takeaways
Azure AI Content Safety EightSeverityLevels returns an integer from 0 to 7, where 0 indicates non-offensive content, catching those who expect a float or a 1-10 scale.
When analyzing a benign image with Azure AI Content Safety using the EightSeverityLevels output type, the API returns an integer 0. This page confirms the correct output format and value for a non-offensive test image.
Common mistake: Choosing 0.0 because of assuming a floating-point output, or misunderstanding the integer-based 0-7 severity scale bounds.
Expert Analysis
Why the Answer Is Correct
Azure AI Content Safety evaluates images for harmful content across categories like Violence, Hate, Sexual, and SelfHarm. When the outputType is set to EightSeverityLevels, the API returns an integer severity score between 0 and 7. A simple circle contains no violent or offensive material, so the service correctly classifies it as non-offensive, returning a severity level of 0.
Why the Other Options Are Wrong
Option B (0.0) is incorrect because the EightSeverityLevels output type returns an integer, not a floating-point number. Option C (7) represents the highest severity level, which would only be returned for extremely offensive or violent content, not a benign circle. Option D (100) is incorrect because the EightSeverityLevels scale only ranges from 0 to 7; 100 is outside the valid bounds of this output type.
Community Comment Notes
Commenters correctly point out that the "EightSeverityLevels" output type ranges from 0 to 7, with 0 indicating non-offensive content. As one user noted, "0 is CORRECT because the image being submitted contains only a simple circle, which does not contain any violent content." Another user confirmed the scale by stating, "The severity levels are from 0 - 7, with 0 being the lowest".
Official Reference
https://learn.microsoft.com/en-us/azure/ai-services/content-safety/quickstart-image
This link opens the official vendor documentation in a new tab.
Exam Strategy
Remember the specific output scales for Azure AI Content Safety. EightSeverityLevels returns integers 0-7, while FourSeverityLevels returns 0, 2, 4, 6. A benign image always returns 0 for any category.
Community Votes
100% of anonymous learners picked answer A. Votes are pick records left by other test-takers — they are not the verified answer.
Discussion 5 contributions
Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.
From public community threads 5
-
3fbc31b 👍 3 Answered AThe severity levels are from 0 - 7, with 0 being the lowest, if any form of offensive material, and 7 being very naughty/offensive. Given this, A is the answer.
-
syupwsh 👍 2 Answered A0 is CORRECT because the image being submitted contains only a simple circle, which does not contain any violent content. When using Azure AI Content Safety with the "Violence" category and the outputType set to "EightSeverityLevels," the service evaluates the image for signs of violence. The "EightSeverityLevels" output ranges from 0 to 7, where 0 indicates "Non-offensive" and 7 represents "Very High" severity. Since a circle is considered harmless, the expected severity score is 0, indicating no detection of violent content. A is correct
-
jafaca 👍 2 Answered A0: Non-offensive 1-2: Very Low 3-4: Low 5-6: Moderate 7-8: High 9-10: Very High
-
mrg998 👍 1In this scenario, you are submitting an image to an Azure AI Content Safety resource for evaluation, specifically looking for violent content using the "EightSeverityLevels" output type. The image you submitted contains a simple circle, which likely doesn't depict any violent or harmful content. Given that the image is non-violent, the expected severity level for violence should be the lowest possible score. In the "EightSeverityLevels" configuration, the scores range from 0 to 7, with 0 representing the least severe and 7 representing the most severe. Thus, the correct output in this case would be: A. 0
-
a8da4af 👍 3 Answered AA is correct.
Reader comments & corrections
No reader corrections yet — spotted an error or have a note? Share it below.