Tutorial 4 Activities#

Pre-readings#

Hopefully you’ve already read the New York Times article “Wrongfully Accused by an Algorithm” (also available on the LMS).

Additionally, you may benefit from reading “Facial recognition: top 7 trends (tech, vendors, use cases)” if you aren’t already familiar with facial recognition technology.

Activities#

Face recognition systems use machine learning algorithms to pick out specific, distinctive details about a person’s face, like distance between the eyes or shape of the chin. These are converted into a mathematical representation and compared to data on other faces collected in a face recognition database e.g. from millions of pictures scraped from websites and social media.

You have already been introduced to facial recognition. In this tutorial, we will explore it in more detail and from the perspective of ethical theory.

Exercise

Is facial recognition ethically justified? In your groups, choose two out of three of the following cases and explore whether they are justified. Include an analysis from the perspectives of utilitarianism and deontology.

  1. Law enforcement

  2. Class attendance at university

  3. City-wide street surveillance for finding missing persons

Emotion or ‘affect’ recognition aims to use machine learning to detect emotions, potentially even very subtle ones—although there is debate about its accuracy—on faces in videos.

Exercise

Is emotion recognition ethically justified? Explore one or more of these cases:

  1. Job interviews

  2. To diagnose depression in patients

  3. In shopping centres to analyse shopper behaviour

Note

Remember to nominate a note taker and post your notes! Here’s a link to this week’s forum discussion.