Shenzhen Micro & Nano Institute (MNI) has licensed the CEVA audio/voice DSP for the development of a multi-mic voice processor targeting a range of intelligent hardware systems including robots, smart appliances and the smart home. “The IoT revolution continues to drive the use of voice and speech processing into new use cases and applications, where microphone array technology is key to deliver the optimal user experience,” said Dr. Feng Xiaoxing, MNI.
“Utilising the CEVA audio/voice DSP, our intelligent voice processor solutions are able to handle even the most complex multi-mic use cases, even in far-field scenarios, in a highly cost-effective and power efficient manner.”
MNI’s advanced voice processing solution supports 8 microphones simultaneously, performing a range of functions to ensure maximum voice clarity and intelligibility in any environment and from any direction.
These functions include echo cancellation, beamforming, noise suppression, far-field voice pickup, sound localisation and speech recognition, allowing the user’s voice to be isolated from any direction.
The complexity of the voice algorithm to handle all of these functions requires a significant amount of DSP processing power, which the CEVA DSP delivers with minimal power consumption.
“By leveraging the superior performance of our audio/voice DSP, MNI can fully exploit the capabilities of its voice processing expertise to create compelling intelligent voice solutions for IoT,” said Moshe Sheier, director of strategic marketing at CEVA. “We’re pleased to add MNI to our long list of customers in China who bring leading-edge CEVA-powered products to meet the demands of an increasingly smarter world.”