New partnership with chipmaker allow image recognition in real time on Android devices
Soon, Android devices can be a lot smarter thanks to a new partnership between Google and Movidius chipmaker that promote lead machine intelligence directly to their mobile devices.
The Movidius specializes in machine vision and has worked with Google in Project Tango, a computer vision platform. Now, through the new collaboration, Google will use the MA2450 of Movidius chip to take deep learning (deep learning) for Android devices.
Deep Learning is one of the ramifications of often applied machine learning to recognize image that uses algorithms to learn at various levels to match other levels of abstraction. Typically, it relies on complex neural networks.
The chip Movidius was built to be extremely efficient in terms of energy, making it suitable to run neural computing networks on smartphones. By deploying its advanced neural computing engines on such chips, Google gives devices the ability to recognize images as faces and street signs in real-time without resorting to connect to Internet and algorithms in the cloud.
These skills could be especially useful for people with visual impairments, for example.
“Our collaboration with Movidius is enabling new product categories to build what people have not seen before,” said Blaise Agϋera y chests, head of Google machine intelligence group.
Financial terms of the deal were not disclosed, nor details of specific plans about products.
“Google is rapidly expanding its business in smartphones in new areas – this is just one of them,” said analyst Jeff Kagan Telecom.
No comments:
Post a Comment