Our Blogs

Google Tensorflow Lite Is A Neural Network Library For Mobile Phones

May 31, 2017
Google Tensorflow Lite Is A Neural Network Library For Mobile Phones
Google has announced the launch of Tensorflow Lite, which is a neural network library for mobile phones. The launch of Tensorflow was announced by Dave Burke, VP of engineering at Google. It is optimized for smartphones. 

As a developer, you can run and execute artificial intelligence applications in real time from within the phone with the help of Tensorflow Lite. Commenting on the launch, Burke disclosed that the library is exclusively designed to be small and also integrate latest trending techniques. 

As of this writing, several artificial intelligence processing occurs on servers of the relevant software as a service provider. The main aim of the Google is to decrease the overload of the processes. 

This technique will help you not only save data but also processing power. Moreover, it will make sure that your crucial data will remain private and can be accessed without an Internet connection.

Touted as the second deep learning tool, the Tensorflow Lite has been used for real-time style transfer purposes. You can easily add art-like filters to your mobile device. To recall, Facebook had announced Caffe2Go framework, which is exclusively developed by the social media giant.

According to Burke, the product team at Google is adding a new functionality to facilitate communication with processors exclusively designed for neural networks. The algorithm can be easily optimized using graphics cards and Googles own Tensor Processing Unit (TPU). If you are unaware, the search engine giant has already deployed neural networks in the Google Translate app.