We all think that Artificial Intelligence is a deep complicated thing, which insist us to think that if we really want to understand it, we have to be serious about it. But Google doesn’t think so. To change the view of people toward AI Google released a AI experiment called Move Mirror in a blog post on July 19. It is a machine learning experiment that matches your poses with images of other people in the same pose.
As humans, it’s easy for us to understand a bodily position and it’s movements, but for machines it’s way tough. Engineers of Move Mirror, develop a computer vision pose estimation model called PoseNet. It is a open-source model. Move Mirror detect the location of 17 different body parts including your shoulders, ankles and hips and identify that in what positions your body joints are.
According to Google’s explainer, it doesn’t take any individual characteristics into account, such as gender, height or body type. Move Mirror watches your move only through your computer’s webcam and with the help of PoseNet and TensorFlow it match your pose against a database having tens of thousands of photos.
After then it compares your pose to more than 80,000 images and finds which ones best mirror your position. Move Mirror then shows you those images next to your own in real time and as you move around, the images you are matched to change. The result is a sort of janky video of different people acting out your movements. You can even make a GIF out of it.
There are video games and imaging device like Microsoft’s 3D , the Kinect which allow users to capture real human movements. But those methods require expensive hardware. With PoseNet Google show that the AI experiment, can also run on a usual hardware. It is impressive how sophisticated machine learning can now be done just in web browsers. Move Mirror may seem like a frivolous, but it is filled with lot’s of fun.
The PoseNet technology is available for developers to experiment with it. It is one of many such Google AI experiments that the company releases from time to time to show there dominance in the field of artificial intelligence.
And if you’re worried about what’s happening with your image when you use it, Google assures that it does not send any of your images to its servers, all the image recognition happens locally and as it is powered by TensorFlow.js all of the pose tracking is done directly in your browser. The technology also doesn’t recognize who is in the image because there is no personal identifiable information associated to pose estimation.
I must also recommend you to try it once. I promise you that it will definitely blow your mind. You can try it for yourself, you just need a webcam, at Google’s experiments page.
Have a great day, Enjoy !
For daily update subscribe to the Techgrabyte newsletter.
Read also, you may like it :