User Tools

Site Tools


sandiego:computer_vision_-_teaching_computers_to_see

Computer Vision: Teaching Computers to See

Author Information

  • Name: Zachary Burns
  • Position and Institution: PhD Student at UCSD, Electrical and Computer Engineering
  • Contact: zburns@ucsd.edu
  • Available for contact for questions and classroom visits: Yes
  • Year contributed: 2020

Categories

  • Grade Level: 9-12
  • Science Domain: Engineering, Technology, and Applications of Science
  • PE Code: HS-ETS1-1, HS-ETS1-2, HS-ETS1-3, HS-ETS1-4

Abstract

Today cameras and computers are everywhere in our society (think cellphones!). Our devices have more processing power than the spaceships that took humans to the moon, however these computers have usually been very bad at doing certain tasks that are very simple for humans. One of these tasks is the ability to recognize and interpret images. Can we teach computers to learn to see like us and what are the applications for science and engineering?

In this lesson students will learn the basics of how a computer uses feature detection and probability to recognize objects. They will explore using an image classification website how well an algorithm is able to do a simple detection problem. At the end of the lesson the students will formulate a science/engineering proposal that uses computer vision and discuss any ethical implications of using their algorithm.

Keywords

  • Computer vision
  • imaging
  • data science
  • machine learning
  • engineering
  • algorithms
  • code

Files

sandiego/computer_vision_-_teaching_computers_to_see.txt · Last modified: 2020/11/01 23:36 (external edit)