More Deep Learning in Cloud Computing

Introduction

Deep learning is becoming increasingly prevalent in cloud computing by providing a number of tools for real-time data analysis. Digital systems will analyze online content directly and expand the metadata processing and exchange network.

Computer Vision

 On August 17, 2017, Box announced the integration of Google’s computer vision technology into its cloud platform. Users can now search images in their files using visual data instead of file names or keywords. According to Box CEO Aaron Levie, as cloud data increase, users will require intelligent ways to organize and understand the content.

Google’s Computer Vision

Computer vision is a technology that enables a cloud system to analyze, detect and recognize visual content in real time. This technology has evolved considerably in recent years with the introduction of deep learning in digital technology fields. To make this form of machine learning more powerful, researchers created artificial neural networks inspired by the way mammalian neurons process and store information. Bio-inspired circuits contain algorithms that can learn to recognize categories of objects. In addition, they can find images with no labelling and perform text recognition on photos or illustrations. Box researchers are also developing tools that will allow customers to develop their own computer vision system.

Computer vision analyzes visual content by isolating main characteristics, such as shape, colour, contrast, and so on. With this process the image of an object can be found in a database or a directory of images by searching the main features of each object.

Box’s computer vision functionality will introduce artificial intelligence and deep learning to a large number of organizations. It will no longer be necessary to manually annotate thousands of images and even directories of very old files, which will not have been tagged, will be searchable. This confirms that deep learning is reinventing cloud computing. Corporations like Google, Amazon and Microsoft continue to offer machine learning features that make the web a powerful work tool due to its pervasiveness, speed, and its huge data management capacity.

Text Recognition

Levie revealed that, in addition to visual and sound analysis, his organization is also interested in text analysis tools. Such tools will be able to perform data searches not with keywords, but with significant entities, for example, they will find all the words that express movement or wisdom.

Stanford University’s Sentiment Analysis technology links adjectives and nouns that express emotions to the sentence structure in order to determine the actual feelings being expressed.

Real-Time Web Content Analysis

Box’s CEO indicated that his team is exploring ways to apply deep learning to the study of customer behaviour. This research would optimize Box service and determine the tasks that would lead to cloud server automation. These tools will one day be able to make decisions and perform specific actions without external intervention.