Homework from Computer Vision – A Happy Home

My NG teacher is simply brilliant! Take a look at what his homework looks like:

“You are going on holiday and have decided to spend a week with your friends. It is a very cosy house with a beautiful surrounding. However, the most important thing is that you have all committed to being happy while you are in the house. Therefore, anyone who wishes to enter the house must demonstrate that they are genuinely happy.

As an AI expert, you must ensure that the ‘Happiness’ rule is strictly adhered to. You will create an algorithm that determines whether a person is happy or not using images from a camera above the front door. The door may only be opened if the person is happy.”

Those who write an algorithm that can recognise a smile in the test data with at least 80% accuracy will pass. Andrew's team managed to achieve this with a simple algorithm at an impressive 95%. That was a challenge. After an hour of hard work, I wrote an algorithm with an accuracy of 97%. Hooray!