Google’s new TensorFlow Lite will soon make Android smarter, giving the smartphone system better AI-dependent features.
(Photo: Photo by Leon Neal/Getty Images) In this photo illustration the Google logo is projected onto a man on August 09, 2017 in London, England. Founded in 1995 by Sergey Brin and Larry Page, Google today makes hundreds of products used by billions of people around the world, from YouTube and Android to Smartbox and Google Search.
This new machine learning framework designed by Google Brain Team is specifically created to work on mobile devices, especially smartphones.
The search engine giant unveiled this new artificial intelligence technology on Wednesday, May 4 for its upcoming Google I/O developer conference, which kicks off on May 11.
Google TensorFlow Lite to make Android smarter
According to computer worldAccording to the latest TensorFlow Lite report, the new TensorFlow Lite is part of the original open-source TensorFlow project.
(Photo: Photo by Leon Neal/Getty Images) In this photo illustration the Google logo seen displayed on a mobile phone and computer screen on August 09, 2017 in London, England. Founded in 1995 by Sergey Brin and Larry Page, Google today makes hundreds of products used by billions of people around the world, from YouTube and Android to Smartbox and Google Search.
“TensorFlow Lite will take advantage of a new neural network API to leverage silicon-specific accelerators, and over time we expect to see digital processing chips purpose-built for neural network inference and training” , said Google.
The tech giant maker added that the new capabilities of TensorFlow Lite will enhance next-generation visual search, on-device speech processing and augmented reality.
For this reason, Google said the new TensorLite Flow framework will make next-generation mobile hardware more advanced. For the past few years, application developers have needed massive computing power just to create advanced machine learning applications.
Now, once the new TensorFlow Lite is widely implemented, application experts should have easier machine learning processes.
Other details of TensorFlow Lite
by TensorFlow official site explained that the new AI-based framework offers different programs that app developers can use to improve their skills. These include in particular the following:
Basic machine learning with TensorFlow (for beginners)
TensorFlow for JavaScript development (for beginners)
Theoretical and advanced machine learning with TensorFlow (for intermediates and experts)
If you want to see more details about these TensorFlow learning programs, you can visit this link.
In addition to these, the new TensorFlow Lite also offers pre-trained models, such as the following:
Object detection (features capable of detecting multiple objects contained in boxes)
Image classification (characteristics for identifying animals, plants, places, people and other objects)
Gesture recognition (features to identify user gestures via webcams)
