Google has overhauled its image recognition technology, which can improve machine vision technology. Notable, the researchers at this America based tech giant has asserted that they have developed an advanced image classification and detection algorithm that is designated as GoogLeNet. Now, it would be possible to recognize a wide range of objects and accurate positions of them with ease.
Hither, talking about GoogLeNet that can quickly figure out things in pictures as well it will label multiple object sizes in a single image. This new discovery of the company manifest itself to determine a motif within or on top of an item.
However, the tech giant has claimed that this new software will be used in any place where it is functional and it is directly transferable to Google products like photo search, image search, YouTube and self-driving cars. Here, the most highlighted advantage of GoogLeNet that it could be useful in manufacturing new devices for visually challenged people.
As per the reports of 9to5google.com, the brand is specially introducing this new technique to serve its self-driving, autonomous car that has already transformed the world of engineering. Articulately, this runabout is highly precise up to a range of 100 meters. It bounces a laser beam on the surface by its LIDAR system and measures the reflection to ascertain the distance and obstacles.
Notably, its public sale is still five years away because the vehicle has been only tested in specific atmosphere and it has limited identification ability such as they can not discriminate Police officers, hurdles on the road. This new quest of the company is yet-to-be examined in different weather and it also can not navigate itself in construction zones.
Here, we can say that Autonomous Car needs to be fast and reliable so that it can refrain itself intelligently from barriers like other motorcars do and the addition of Google's new software will surely arrive as a profitable thing for it.