Voice recognition software market to become more diverse: Companies and Markets

January 30, 2015 - 

A new report published by Companies and Markets predicts that the global voice recognition market will become more diverse and see significant growth as the accuracy of the software improves.

The “Global Voice Recognition Market: New Insights” report states that the global voice recognition market is projected to reach a value of $133 billion by 2017 as tech giants like Google, Facebook, Apple and Microsoft continue to invest in voice recognition software.

For instance, Facebook recently acquired Wit.ai software, which enables developers to create an API for building voice-activated interfaces.

Additionally, companies like Nuance Communications, Microsoft Corp. and ValidSoft Ltd, along with Agnitio, Advanced Voice Recognition Systems, Auraya Systems, LumenVox, Raytheon BBN Technologies, Voice Biometrics Group, Voice Trust and Voxeo are all recognized as global voice recognition market leaders.

The report emphasizes the main challenge of the global voice recognition market being the lack of accuracy in voice recognition systems.

The analyst authoring the report explains that in order to work, voice recognition applications must be highly sound-sensitive. Taking this into account, the accuracy of voice recognition applications are greatly impacted when used in noisy everyday environments.

Market experts believe that the sound and visual analysis components of speech recognition solutions must be fully amalgamated in order to improve the accuracy of voice recognition software in the future, says the report.

The report also highlights work being done by researchers in developing lip-reading apps, such as Mu’tah University’s Ahmad Hassanat. Additionally, Yasuhiro Oikawa of Waseda University in Japan has been developing methods of monitoring the movements of the speaker’s throat in order to accurately interpret speech.

Developing these kinds of visual technologies will improve the accuracy of voice recognition software by differentiating between similar sounding words and reducing the disruptions caused by external noise, said the report.

Leave a Comment


About Stephen Mayhew

Stephen Mayhew is the publisher and co-founder of Biometrics Research Group, Inc.. Connect with Stephen on LinkindIn.