Google introduced Quick024 Archives Draw! last November at our productivity's great expense. We spent hours in the office toying around with the image maker, resulting in some next-level sketches and a firm lecture from our bosses about "proper time management," whatever that means.
The tool wasn't just meant to help us waste time, though -- Quick, Draw! was actually a clever crowdsourcing effort from Google's A.I. Experiments team to train its neural network. As people played the game, the system learned how to identify exactly what they were sketching more accurately.
Google did us all a solid and released the data gathered from the program to give us all a look at how 15 million people have used Quick, Draw! since November. The cache has over 50 million drawings spread out over 345 different categories, so there's clearly been plenty of time spent doodling away online -- which is great for the AI, as we saw at the company's recent I/O developers conference. By releasing the dataset, Google hopes to give everyone a chance to use the crowdfunded sketches, especially developers trying to train neural networks of their own.
The data might be for aimed at developers -- but that doesn't mean we can't all enjoy it. With all the drawings released, the real fun is in scrolling through the endless rows of user-drawn sketches as they pop up in real-time.
Here's a few animal sketches in GIF form -- but you should totally check out the full set, too.
Nothing is better than the line. It might surprise you, but there more than just one way to drag your finger (or mouse) from point A to B.
If you're a developer, researcher, or anyone else with more than just a casual interest in the work, you can check out the full dataset on Github. That way, you can actually help make all of that time we wasted worthwhile.
Topics Artificial Intelligence Google
(Editor: {typename type="name"/})
Impact of Temperature on Intel CPU Performance
Electric vertical takeoff and landing craft completes first flight in Shenzhen · TechNode
China’s Fudan University launches AI app for visually impaired individuals · TechNode
Meizu shifts to AI, ceases traditional smartphone development · TechNode
10 Free Steam Games Worth Playing
Li Auto shares plunge on bleak order forecast for first BEV · TechNode
NIO battery swap stations facilitate peak
Nvidia Pascal Goes Mobile: GeForce GTX 1080, 1070 & 1060 Preview
Stellantis reportedly considers manufacturing Leapmotor EVs in Italy · TechNode
Free Rita's Italian Ice: How to get free Italian Ice on March 20
Huawei overtakes Apple in China's Q4 2023 tablet market · TechNode
接受PR>=1、BR>=1,流量相当,内容相关类链接。