Applied Artificial Intelligence: Speech Recognition Systems

Explore AI-powered technology

This course is part of the Advanced and Applied AI on Microsoft Azure ExpertTrack, helping you develop AI and machine learning skills and prepare you for the relevant Microsoft microcredentials.

This course will teach you the fundamentals of the components of a modern Automatic Speech Recognition (ASR) system. You’ll then put this knowledge into practice by building your own speech recognition system almost entirely out of Python code, a powerful tool used across AI and data science practices.

Delve into Automatic Speech Recognition

When a human speaks a word, they cause their voice to make time-varying patterns of sounds, and waves of pressure that spread through the air.

During this course, you’ll understand how these sounds are captured by a sensor, turned into a sequence of numbers and how an automatic speech recognition system converts this into a textural representation of what was said.

You’ll delve into the components of ASR as well as the fundamental theory and background of speech recognition.

Build your own speech recognition system

You’ll identify the different models and problems when designing Speech Recognition Systems as you build your own. This hands-on approach will help you identify the different components of speech decoding and demonstrate your knowledge and understanding of techniques such as Advanced Acoustic Modelling.

During each lab, you’ll build a different functioning block of the system and by the end of the course, you will have built a speech recognition system almost entirely out of Python code, a powerful tool that you can use across your data science practices.

This course is aimed at anyone with an understanding of data analysis who has created models using machine learning. Individuals who will benefit from this course include:

• Data Analysts
• Machine Learning Engineers
• Deep Learning Engineers
• Many more professionals involved in the development of AI-based technologies.

• Course Introduction
• About this Course
• Introduction to Fundamental Theory
• CloudSwyft Hands-On Lab 1
• Speech Signal Processing
• CloudSwyft Hands-On Lab 2
• Wrapping Up the Week
• Acoustic and Language Modeling
• Acoustic Modelling
• Deep Neural Network Acoustic Models
• CloudSwyft Hands-On Lab 3
• Language Modelling
• Language Model Evaluation and Operations
• Wrapping up the week
• Speech Decoding and Advanced Acoustic Modeling Techniques
• Week 3 Welcome and CloudSwyft Hands-On Lab 4
• Speech Decoding
• WFST, Acceptors and Graph Composition
• CloudSwyft Hands-On Lab 5
• Advanced Acoustic Modeling Techniques | Improved Objective Functions
• Course Completion