Voice biometrics award and upgrade announced as Australian service grows in popularity
Speech Technology Center’s speech recognition technology has been globally recognized at the international CHiME Speech Separation and Recognition Challenge (CHiME-6) for its ability to recognize English speech from multiple microphones in a natural environment, the company said.
At CHiME-5, participating companies had to identify spontaneous speech from several speakers when speech overlaps with noise. Unlike the CHiME-5 challenge, CHiME-6 involved working with non-segmented speech overlapping up to 20 percent. Speech Technology Center focused on this task, and outperformed competitors, according to the announcement.
Detection was performed on recordings from 20 genuine parties and involved the simultaneous speech of 2 to 4 people mixed with different housekeeping sounds. The challenged involved developing a recognition system that would first listen and then create a full transcript of the recordings.
Speech Technology Center developed an algorithm that separated speech segments for each speaker and applied neural networks to tell the difference between speakers. The company competed against Toshiba, Johns Hopkins University (USA), University of Science and Technology of China, Brno University of Technology (Czech Republic) and others.
“This year, at CHiME-6, for the first time in history, we faced the most challenging task: working with non-segmented speech. The ability to precisely recognize the speech by different speakers, even when interrupted by noise, allows these services to outgrow the innovation status and enter everyday use, assisting businesses and making our lives easier,” said Dmitriy Dyrmovskiy, CEO of Speech Technology Center, in a prepared statement.
“High-quality processing of non-segmented speech will allow, for example, conducting competent logging of meetings where several speakers speak at once, while the intelligent speech analytics will help automate the work of call centers by recognizing spontaneous speech, classifying voice calls, checking compliance with the script, assessing customer satisfaction and quality of dialogue, thus significantly optimizing the workflow of modern call centers in retail, e-commerce and telecom,” Dyrmovskiy added.
LumenVox upgrades biometric speech recognition solution
LumenVox has released an updated version of its biometric speech recognition solutions which includes a number of enhancements to make speech applications secure and easy to deploy, the company announced.
Top upgrades include support for additional connections, testing and enhanced security with encrypted audio packet transmission, additional short utterance transcription language support, up to date operating system support for optimal security, performance, development and streamlined installation, added flexibility regarding MRCP properties, and extended dashboard diagnostic capabilities for TLS protocol functionality and testing.
LumenVox has recently been honored by Speech Technology Magazine with three People’s Choice Awards for its biometrics, speech recognition and text-to-speech technology.
Services Australia enrolls over 1.2 million people with Nuance voice biometric authentication
Services Australia has a voiceprint database of over 1.2 million people who signed up to access government services following upgrades to the voice biometric authentication service provided by Nuance , writes IT News.
“As at 19 March 2020, 1.2 million customers have enrolled a voiceprint with Services Australia,” it said according to the publication. “Services Australia made it easier for customers to register their voiceprint; this has resulted in customers registering voiceprints in increasing numbers.”
The highest number of welfare applicants was registered during the coronavirus lockdown. Citizens enrolled their voiceprints by saying “in Australia my voice identifies me” to confirm identity when contacting the call center.
“A biometric voiceprint is the strongest level of security available to protect our customers’ information over the phone,” Services Australia said in a prepared statement.
Nuance voice biometrics was also chosen by the Australian Taxation Office in 2014, which has already enrolled more than 4.2 million voiceprints that can be used to access the contact center and mobile app, after reaching 3.4 million in early 2018.