Microsoft Azure AI Fundamentals (AI-900)Describe features of computer vision workloads on AzureMedium

A manufacturing plant wants to implement a quality control system that automatically inspects finished products for surface defects, such as scratches or dents. The defects are unique to their product line and require very specific visual identification. Which Azure AI Computer Vision tool or service would be most appropriate for training a model to detect these specific defects?

  1. AAzure Computer Vision API
  2. BAzure Face Service
  3. CAzure Custom Vision
  4. DAzure Form Recognizer
Show answer & explanation

Correct answer: C. Azure Custom Vision

Azure Custom Vision allows users to train custom image classification and object detection models for specific visual domains, which is ideal for detecting unique, product-specific defects that a pre-trained API might not recognize.

Why the other options are wrong

  • A. Azure Computer Vision API provides pre-trained models for general tasks like tagging and object detection, but it might not be specialized enough for unique product defects.
  • B. Azure Face Service is for face-related tasks (detection, recognition, analysis), not general object or defect detection.
  • D. Azure Form Recognizer extracts data from forms and documents, which is irrelevant for visual surface defect inspection.

Azure Custom Vision

An Azure AI service that allows developers to build, deploy, and improve custom image classification and object detection models for specific use cases.

  • Requires training with your own labeled images.
  • Ideal for niche or specialized computer vision tasks not covered by general pre-trained models.
  • Supports both classification (what is in the image) and object detection (where are specific things in the image).

Memory trick: Custom Vision: Train your eye for unique targets.

More Describe features of computer vision workloads on Azure questions