Introduction to audio classification with TensorFlow

  • Module
  • 5 Units
Beginner
Data Scientist
Developer
Student
Azure Machine Learning

Learn how audio is represented as tensors, how to convert waveforms into spectrograms, and how to use TensorFlow to build a simple keyword classifier.

Learning objectives

In this module you will:

  • Describe how sample rate, amplitude, channels, and waveforms represent audio data.
  • Convert audio waveforms into spectrogram tensors for model training.
  • Build and evaluate a binary TensorFlow keyword classification model that recognizes "yes" and "no".

Prerequisites

  • Basic Python knowledge
  • Basic understanding of machine learning
  • A Python environment that supports TensorFlow 2.10 or later, with TensorFlow and Matplotlib installed
Start