Wednesday 4 January 2017



शुभ दिन धूमकेतु! (Good day Comets)


     At our last meeting we had a special guest from Google Deep Mind in London, Dr Tim Lillicrap. Dr. Lillicrap taught us how to train our convnets to recognize animals such as wildebeests. Dr. Lillicrap also showed us a project his team at Google did where the taught a convnet to go around a virtual house and collect apples. The word of the week is Apples. Dr. Lillicrap's team  also trained a convnet to play GO, a game popular in Asia, better than any other human. You can read about that here.
Tim Lillicrap sharing his knowledge with us




 We also contacted Roy Chan, a computer science major at Stanford University who wrote a paper on training a convolutional network called YOLO (You Only Look Once) to recognize baked goods. We asked him what software he used and how long he would think it would take us to get our project off the ground (ha ha, pun intended 😆) and he told us it would probably take 3-6 months. To read Roy Chan's paper click here. ( Bonus points if you can say the bakery he got his baked goods from.

Also the Ontario East Provincial Championship is in 10 sleeps so everybody need to review their lines and check Slack for any homework. If you are not signed up for Slack, get your parents to check their email (or check your own) so you can get signed up.