Homework: Street-Level Images for Inspiration

Background

To ideate a machine vision-based urban AI system, and to plan the collection of a proper image dataset, you need a starting point. This activity is about going out and collecting a first sample of street-level images that you can use in an ideation exercise next week.

Learning Objectives

After completing this activity you will have gained some understanding of what a camera on street level can and cannot capture. You will also have spent some time thinking about what kinds of issues can be detected using street-level image data.

Instructions

• Go outside and bring a camera. The one on your smart phone is perfectly fine.
• Walk around your neighbourhood and look around. Take photos of things that catch your eye.
• At this point, don’t worry too much about how useful the photos are. For example, they do not need to be concretely related to an urban issue yet. The purpose is to collect an inspiring set of images. Go for variety, not volume.
• Spend at least about 30 minutes walking around and taking photos.
• When you are done, download the photos from your camera or phone, and upload them to a cloud service so that you can easily share them during next week’s session.

Product

The result of this activity is a set of photos, at least roughly 15, showing a variety of street-level scenes. They are stored in a cloud service for easy sharing.

Follow-up

In next week’s class, we will have a workshop where you will show the images to your group, discuss them, and use them as inspiration for ideating an urban issue to work on and a proper image dataset to collect.