Artificial Intelligence - AI-900 - Vision Workloads


Posted by Willy-Peter Schaub on Fri 27 September 2024

"Vision workloads in AI refer to tasks that involve processing and analyzing visual data, such as images and videos." - ChatGPT GPT-4o

alert These are my living personal study notes. Use them at your own risk!

vision-workloads

Notes

  • Custom Vision service has a model you can build and train for classification or object detection (class name, probability score, bounding box).
  • Computer Vision detects faces, Face Service finds and recognizes similar faces.
  • Face Detection helps detect and recognize people's faces.
  • Form Recognizer service has a custom model users can train, or a pre-built receipt model trained to work with receipts.
  • Image Analysis helps extract information from the images, tag them, and create descriptive image summary.
  • Image Classification classifies objects based on their content to answer the "what is the image?" question.
  • Semantic Segmentation classifies pixels that belong to objects (detects) and highlights the object boundaries.

You perused:

Enjoy other learning notes:

AI-900 Quick reference Poster: