Handwriting Recognition

Handwriting recognition is a special case of image classification. It is often used as an introductory example to learn about how a computer can learn to "see" and recognize concepts in images.

A common task for AI is handwriting recognition. This is a special task in the broader field of computer vision. Using this example, we try to understand how a computer can learn to "see".

Key Takeaways

• Computers using machine learning algorithms have surpassed human performance in image classification and are now widely used for many tasks.

• To a computer, an image is a large list of numbers. The task is to learn the patterns that relate to concepts we want to recognize in an image.

• To teach a computer to recognize handwritten digits, we need many examples (training data) with the correct answers (label).

Slides

You can open the slides in a new window or download them as PDF.

Video