Perfect Speech Intelligibility
AI Cleans Voice from Noise
OverviewIn emergency situations like accidents on the road or in a factory, every second counts. If background noise or echo effects distort the spoken word or make it completely unintelligible, people’s health or even lives are at stake.
Our AI Cleans Voice from Noise technology is the game-changing solution to the background noise problem. Using an AI-based technique known as ‘deep neural networks’, the solution is able to eliminate background noise and enhance the voice signal to an amazing quality level. The result is similar to direct face-to-face communication in a noise-free environment.
Trusted Voice
To actually clean voice from noise, we worked in close research cooperation with the Institute for Electronic Music and Acoustics (IEM) at the University of Music and Performing Arts Graz. With joint forces we have been able to significantly optimize the signal processing of our Intercom module im3 and the Intercom station od1. The result is a voice profile that shines with an unprecedented level of clarity and intelligibility, much like a high-quality transmission in a noiseless environment.
How it works
The secret to this lies in so-called deep neural networks. Deep Neural Networks are like interconnected networks of artificial brain cells that learn from data to solve various kinds of problems, similar to how our brains learn from experiences to make decisions. They ensure that the voice signals during calls are cleaned completely from ambient noise and echoes during transmission.
We have always been proud of our high-quality voice transmission. With the trained neural networks we have now raised the quality mark even higher for our Trusted Voice. Any residual noise is suppressed to near perfection, as the result of a comprehensive training process using standardized machine learning algorithms. By feeding them countless hours of audio data, they are systematically ‘taught’ the desired recognition capability. The neural network now reacts with extreme precision to the many different challenges of everyday life – things like different voices and a wide spectrum of ambient noise in a variety of situations.
We encourage you to hear the exceptional result for yourself:
