Vision Recognition

Classification of an image

Vision is a key part in our day to day lives. Almost everything we do requires us to use our sense of vision more than our other senses. Common activities such as driving and sports. We use vision to identify and understand our world. Tell different colours apart, name different objects and navigate our way through the world. The same way an AI can be taught to identify and understand our world.

In this workshop, you will investigate an AI robot and figure out how it thinks. You will use our vision recognition AI to take picture of everyday objects around you and see what the AI robot thinks it is. The robot will tell you what kind of object it thinks the picture is and give you some classifications. If there is a classification you don’t agree with, you can select misclassified and send your result to an image library.

Misclassification of images

Your friends will also be sending their misclassified images to the library as well and it will be everyone’s job to find the correct classification of the image. You will also have some challenges to try and trick the AI into misclassifying an object or obscure an object. This will help you figure out how the AI robot thinks!

Vision recognition can in fact enhance our capabilities and we can accomplish so much more! Autonomous cars, which was once considered a science fiction. can now be brought to life with the help of vision recognition. By training an AI with image data, we can teach it to recognize objects, such as a stop sign, green light, red light etc. which can then be used to control a car. Autonomous cars are just one of many cool things that we are able to do with vision recognition, there are many other endless possibilities with vision AI!