Master Computer Vision in Azure: A Guide to AI-900 Certification

📕 Table of Contents:

1. Introduction
2. Provisioning Cognitive Services in Azure
3. Computer Vision Service
   • Standalone Service
   • General Cognitive Services
4. Training Models with Custom Vision Service
   • Image Classification
   • Object Detection
5. Analyzing Faces with Face Service
   • Face Detection
   • Facial Analysis
   • Facial Recognition
6. Reading Text with Computer Vision Service
   • Optical Character Recognition (OCR)
   • Use Cases of OCR
7. Analyzing forms with Form Recognizer Service
8. Conclusion

📘 Article:

Introduction

Welcome back! In this AI 900 certification Course, we will be focusing on Microsoft Azure AI Fundamentals, specifically diving into computer vision in Azure. In this article, we will explore the various cognitive services offered by Azure and how they can be used for computer vision tasks.

Provisioning Cognitive Services in Azure

To make use of cognitive services, you need to provision a resource in your Azure subscription. There are two options available: standalone service-specific resources and general cognitive services resources. Both options offer their own advantages. Standalone resources allow for independent management of cost and access, while general cognitive services resources provide centralized management for all AI services.

Computer Vision Service

Azure offers a pre-trained computer vision model that enables the analysis of images. To use this service, you need to create a computer vision resource in your Azure subscription. Alternatively, you can choose to create a general cognitive services resource, which includes computer vision along with other cognitive services.

Training Models with Custom Vision Service

The Custom Vision service in Azure allows you to train custom models for image classification or object detection. Image classification involves training a model using categorized images, while object detection involves recognizing and locating specific objects in an image. To create a custom vision solution, you will need a custom vision resource and a cognitive services resource in your Azure subscription.

Analyzing Faces with Face Service

Face detection and analysis is an area of artificial intelligence that focuses on locating and analyzing human faces in images or videos. The Face service in Azure provides capabilities such as face detection, facial analysis, and facial recognition. Face detection involves identifying regions in an image that contain human faces, while facial analysis can provide information about facial landmarks, age, and emotional state. Facial recognition allows for the identification of known individuals.

Reading Text with Computer Vision Service

Computer vision capabilities can be utilized to process written or printed text. Optical Character Recognition (OCR) is a fundamental aspect of text processing, where models are trained to recognize individual shapes as letters, numerals, or punctuations. This enables the reading of text in images or documents. OCR has various use cases, such as note-taking, digitizing forms, or scanning checks for banking purposes.

Analyzing Forms with Form Recognizer Service

Many organizations face the challenge of processing receipt or invoice data. Azure's Form Recognizer service offers an AI solution that not only extracts text data from receipts but also intelligently interprets the information contained in them. This alleviates the need for manual review and enhances the efficiency of processing large volumes of receipts and invoices.

Conclusion

In this article, we have explored the different cognitive services offered by Azure for computer vision tasks. We discussed the provisioning of cognitive services in Azure, training models with the Custom Vision service, analyzing faces with the Face service, reading text with the Computer Vision service using OCR, and analyzing forms with the Form Recognizer service. The next module will delve into Natural Language Processing (NLP), starting with an introduction to NLP.

🔍 Highlights:

• Azure provides a range of cognitive services for computer vision tasks.
• Standalone or general cognitive services resources can be provisioned in Azure.
• The Custom Vision service enables the training of custom models for image classification and object detection.
• The Face service offers capabilities such as face detection, facial analysis, and facial recognition.
• Optical Character Recognition (OCR) allows for the reading of text in images or documents.
• The Form Recognizer service intelligently interprets information from receipts and invoices.

❓ FAQ:

Q: Can I use multiple cognitive services under one resource? A: Yes, you can use a general cognitive services resource that includes multiple services, such as computer vision, text analysis, and translation.

Q: What is the purpose of face detection in computer vision? A: Face detection is used to identify regions in an image that contain human faces, typically by returning bounding box coordinates.

Q: How does OCR benefit organizations? A: OCR enables organizations to process written or printed text, eliminating the need for manual data entry and facilitating tasks such as digitizing forms or scanning checks.

Q: Can the Form Recognizer service extract specific information from receipts or invoices? A: Yes, the Form Recognizer service can extract text data from receipts or invoices and intelligently interpret the information contained in them.

Q: What will be covered in the next module? A: The next module will focus on Natural Language Processing (NLP), starting with an introduction to NLP techniques and concepts.