Date: 03/11/2016

XMOS Improving voice control on embedded systems through partnership

Voice interface and recognition is becoming important for microcontroller chips to support voice controlled IoT devices or any such devices which need to be instructed through voice commands.

The microcontroller chip vendor XMOS is offering support in its chips for processing voice by partnering with companies who have expertise in speech recognition/voice processing. XMOS is partnering with a company called Setem technologies to improve speech recognition from far fields.

"Combining powerful XMOS multicore processors with Setem's far-field pre-processing technology will create an embedded voice solution that improves the quality of far-field speech recognition and communications," said Jochen Meissner, CEO of Setem Technologies. "For OEMs, our joint solution will shorten the time-to-market for IoT products that require precise voice control."

Setem uses its own platform called 3S pre-processing for improving the speech recognition for far field voices. This feature enables appliances such as smart-home, connected car and Internet Of Things applications can be voice controlled from far away.

"Setem's signal separating technology adds the precision of voice localization to XMOS, which is essential in far-field applications," said Mark Lippett, CEO of XMOS. "Our product development and marketing teams are working together to create a best-in-class voice solution that will drive the adoption of voice user interfaces in IoT products."
The Setems 3s platform uses patented source signal technologies to isolate and separate and enhance and de-reverb signals to increase the ability to decode the speech accurately. The XMOS xCORE-200 chip supports to handle such audio related processing.

XMOS is also partnering with another company called Sensory to supports its customers to use Sensory's TrulyHandsfree voice control technology in the XMOS xCORE-VOICE far-field voice capture processing platform.

TrulyHandsfree is a voice keyword detection software from Sensory, where it provides accurate sensing even in noisy environments.

The xCORE-VOICE platform for XMOS supports microphone aggregation combined with sophisticated voice processing and also flexible interface so that the speech recognition can be made automatically.

Todd Mozer, CEO of Sensory said "XMOS is driving new levels of integration and performance in the voice processing market. The combination of its hardware and software solution platform with our TrulyHandsfree technology delivers a potent combination into this rapidly expanding voice-enabled market."

Mark Lippett, CEO of XMOS said "We are delighted to partner with Sensory to integrate its industry-leading keyword detection technology as part of the xCORE-VOICE solution. We expect our partnership to help drive the adoption of voice interfaces across a wide range of consumer products."