Google working on Image Recognition

Everything in today’s world is extending to a larger scale. And when it comes to technology, well there are no limits of exploring it and introducing a larger scale of its various forms.

Nowadays “Image Recognition” is a hot and challenging research topic which is being spotted by all the big firms like Google, Microsoft, Facebook etc. They are using this technology same like Facebook detects the faces of the people to tag them automatically using image recognition technology.

Definitely if we talk about the computer world – Graphics is a very important and attracting part and graphics is something which includes Image Recognition itself and then it gets extended to neural networks plus image recognition.

Now this is the nerve which is grabbed by Google and now Google working on Image Recognition on a larger scale to improve the Image Recognition advantages and tasks. Google is planning to launch a software which can detect the angle of the object even with respect to the room very accurately.

READ ALSO: Weapons of intelligence or death?

One of the biggest challenges in the field of computer vision technology is ImageNet Large-Scale Visual Recognition Challenge i.e. ILSVRC. Image can be recognized easily nowadays but what if the Image can be understood as well? Like it’s movement angle, its location, expressions, outfit and everything.

Google has hired a special team for this visual and neural networks research which is named as GoogLeNet. The team will be working in three decomposed steps of Image Recognition i.e. Classification of the image which will include all about the physical outlook of the image, second is Localization of the image that is where the image or object is located with a proper and accurate angle and the third step is Detection which is same as the Classification thing but where the difference lies is it’s going to determine the unique features of that particular object with respect to other objects.

The above three steps seems to be just simple tasks to be performed but people this is a pretty hard and challenging task to be accomplished with 100 percent efficiency giving accurate results.

This particular software or GoogLeNet model is going to work for a live camera of your gadget, say a phone, computer or other devices. So, ultimately it’s like having a camera with artificial brain. In that case, this is something which relates to Artificial Intelligence as well.

READ ALSO: Future robots will be controlled by thoughts

So, let’s just wait for this exciting Robo-camera making kind of product by none other than Googlers.

Join the discussion

Leave a Reply

Your email address will not be published. Required fields are marked *