In art class, the students would get to see a model (a flower vase, a basket of fruits, vegetables, or a semi-nude or nude guy/girl on a high stool) for only a few minutes. And then the model would be gone, and the students would get busy drawing/watercoloring/etc.. of the scene from their mind's impression. What if one is given a word, and 20 seconds to interpret that word in a diagram. Email friend sent me a link that offers this fun thing to do. Click on the link below... and start the game. You will see a word/phrase, and have 20 seconds to do a finger diagram on screen. The computer will then try to recognize what you drew, and if it recognized the descriptor word or phrase, you'd score a point. Quick, Draw! Can a neural network learn to recognize doodling? Help teach it by adding your drawings to the world’s largest doodling data set, shared publicly to help with machine learning research. post your score in this thread...
I got 5 out of six. I was asked to draw a "pool" as the first one. While I started with the diving board, it said a diving board, but it took me 22 or so seconds to complete the pool. The guesses by the computer (neural networks) were all interesting... The rest of the five were easily guessed : Ocean, mountain, hand, dumbbell, golf club.
There has been a growing no of data viz projects based on the google doodle data. Here is one that caught my attention recently. How Long Does it Take to (Quick) Draw a Dog?
"Windmill" was one of the asked-for doodles... When the scatter-brained look at that (yoohoo @Blinky ) they'd think of various things.... including Sancho Panza and the Don.... and lose a few seconds giggling over that.... as in, [I added the windmill over a borrowed Quijoté]
Thank you @Amulet ! I've bookmarked this page. will be playing this everyday! And interesting information to go with it also! I was lucky the first time!
Haha thank you so much. I had fun with this. This motivates me to get back to Andrew Ng's intro course on machine learning.