Top Ad unit 728 × 90

Recent News


Apple Looking to Enhance Speech Recognition through Geolocation

Apple seems to be working to find new ways to enhance speech recognition on its devices.

Now it looks like the company is willing to achieve this through geolocation, as well as localized language models.

Siri is one of the features that can currently be found in iOS devices, when it comes to voice recognition.

However, Apple may not be that happy with the way this service functions now. So, the company is searching for ways to improve it.
Speech Recognition through Geolocation

The main problem that the famous manufacturer has tracked is the lack of accurate representation of some local dialects and languages, which can be an issue of this service.

 The solution that Apple seems to have found is integrating location data with language modeling, with the purpose to create a hybrid system which may prove more efficiency when it comes to understanding a larger variety of languages.

What the new Apple invention aims to do is to hybridize local language models, analyzing location and speech input, then merging them with other models. In global models, there are included the general language properties, which are commonly used by native speakers.

The local language model can be determined by geography. Location data will be obtained via GPS, but the user can manually enter the location, too.

The need for the company to work at such a new system comes from Apple’s attempt to promote the iPhone in a wider range of countries.

Apple seems to believe that creating a system that includes the speaker’s location can actually improve the likelihood of a command of a question to be understood. So, from this point of view, Siri may be able to better understand speech patterns when geolocation data is introduced in the system.

 It is yet unclear if Apple uses this new technology or plans on introducing it in its latest devices. Moreover, it is unknown if the company plans on introducing it to its entire iOS product line. Currently, Apple devices have speech recognition features.

However, the location-based recognition technology was only presented for the first time in 2012 and its inventor appears to be Hong M. Chen.
Apple Looking to Enhance Speech Recognition through Geolocation Reviewed by Mike Beasley on 2:33 PM Rating: 5
All Rights Reserved by Madd Apple News © 2014 - 2015
Powered By Blogger, Designed by Sweetheme



Email *

Message *

Powered by Blogger.