AI-102 Question 98
Select 3You are tasked with implementing an image moderation solution for a social media platform using Azure AI Content Safety. The solution must detect and flag images containing adult content or violence. Additionally, the solution should be able to integrate with your existing Azure-based application where flagged images are sent for further review. Which steps should you take to implement this solution?
- A
Use the Azure AI Content Safety API to analyze the images for harmful content.
- B
Enable the Azure AI Content Safety API to directly delete flagged images without additional review.
- C
Configure the Azure AI Content Safety API to return confidence scores for detected content categories.
- D
Integrate the Azure AI Content Safety API with your existing application to route flagged content to a review pipeline.
- E
Manually train a custom image moderation model using Azure Machine Learning and replace the Azure AI Content Safety API.
Show answer and explanation
Correct answers: A, C, D
Explanation
To implement an image moderation solution with Azure AI Content Safety, you should use the API to analyze images and configure it to return confidence scores for harmful content. Integrating the API with your application enables flagged images to be sent to a review pipeline for further processing. Manually training a custom model or allowing the API to directly delete flagged content is not required or recommended in this case.
- A. Correct.
Correct. The Azure AI Content Safety API is designed to detect harmful content, including adult content and violence, making it suitable for this scenario.
- B. Incorrect.
Incorrect. The Azure AI Content Safety API does not directly delete content. It provides insights and flags for you to handle flagged images as per your business logic.
- C. Correct.
Correct. The API provides confidence scores for detected content categories, which is essential for determining the severity of flagged content.
- D. Correct.
Correct. Integrating the API with your application allows you to route flagged content to a review pipeline for further analysis or action.
- E. Incorrect.
Incorrect. Manually training a custom model is unnecessary because Azure AI Content Safety already provides pre-trained models for content moderation.