In the mid-80s, Buick presented a combined “entertainment” and “information” system, with a 9-inch screen in its all-new Rivieras luxury sedan. But, unfortunately, this futuristic luxury failed to stick.
Now, two decades later, in-car infotainment is one of the most important factors to customers in new car purchases.
Today, manufacturers equip economy cars with advanced infotainment systems that enable multiple voice-based tasks.
Voice interaction is one of the most important functions performed by infotainment systems in connected cars. We use voice interactions to make hands-free phone calls and perform tasks like getting directions, booking reservations, or playing music.
We will eventually use voice interactions to activate and even drive our cars. Therefore, in-car use is a crucial driver of voice assistant adoption.
Considering hands-free communication, speech quality in many infotainment systems is acceptable. Moreover, the vast majority, if not all, of the systems comply with the International Telecommunication Union (ITU) P.1100 standard, even in the presence of road noise.
However, as noted in a previous article, “Speech Recognition In Cars - An Experimental VUI Study,” road, wind, and other background noise reduces speech recognition quality in cars, in turn negatively impacting a voice assistant’s performance.
Moreover, the presence of multiple people talking can also impact speech recognition quality.
In collaboration with HEAD Acoustics, GmbH, a global leader in audio and vibration analysis, we analyzed hands-free telephony speech quality in a car, focusing on use cases where multiple people are speaking simultaneously.
We also tested speech recognition rates and analyzed speech quality at speeds from 0 km/h to 120 km/h. We conducted the study using the existing infotainment system of a Renault Megane GRAND Tour.
The two-phased study shows that Kardome’s location-based source separation technology surpasses existing in-car hands-free telephony (HFT) and speech recognition systems.
Sign up below to read the full study.
Download the Speech Quality Study:
Your submission has been received!