NXP & Amazon talk up voice-enabled devices

NXP & Amazon talk up voice-enabled devices

NXP & Amazon talk up voice-enabled devices
Amelia somehow felt strangely out of focus once she had installed her new Amazon Echo voice recognition smartspeaker. Image Credit: Amazon

Automotive semiconductor and cybersecurity specialist NXP Semiconductors has released its ‘reference platform’ for Amazon’s Alexa artificial intelligence-based personal assistant technology.

Dutch headquartered NXP has made this move to provide a connection point between its own technology and Amazon’s far-field voice recognition technology and Alexa Voice Service (AVS).

But what does any of this mean and why is it important?  

Buzz-phrase demystifier

A ‘reference platform’ is a technology blueprint or specification  intended to be copied by software application developers and systems architects based (in this case) upon internal knowledge of a chip maker’s ‘chipset’. It’s like a map and a how-to guide, rolled into one.

Amazon’s ‘far-field’ voice recognition technology is so named because the company’s ‘Echo’ smart speaker unit is built with seven internal microphones. This means it can pick up sounds from a 360-degree scanning perspective, even where there is background noise. Amazon has previously also offered the ‘reference design’ for its microphone array (with the custom software algorithms that power it) to outside hardware developers through its Alexa Voice Service program.

Come together, over IoT

The coming together of NXP and Amazon’s technologies could help simplify the creation of new voice-controlled IoT devices.

What NXP has done is to encapsulate ‘a complete system’ in the form of this reference platform for Amazon Alexa on its i.MX processors. This contains Amazon’s 7-microphone array design, its far-field audio processing technology and the AVS client, in order to enable developers to create their own devices with Alexa.

NXP’s i.MX applications processors have IoT-centricity on three levels. The i.MX 6 series is intended for general purpose applications (general-use apps that work, for example, in automative displays inside a car). The i.MX 7 series is for low-power applications (as used in core IoT sensors and other related component level apps). Finally, the advanced i.MX 8 series is for richer, interactive high-performance multimedia and audio.    

“Integrating high-quality audio processing has made the development of advanced voice-enabled devices lengthy and complicated,” said Geoff Lees, senior vice president and general manager of microcontrollers at NXP.

“NXP’s reference platform for Amazon Alexa is the definitive solution to this problem. This reference design integrates Amazon’s proven far-field voice recognition technology and our popular i.MX development platform to enable the creation of high-performance voice-enabled devices with Alexa and reduce time to market.”

Read more: Smart home product manufacturers must target customers in different ways

Beamforming beanfeast

The NXP Reference Platform for Amazon Alexa features the same far-field technology developed by Amazon for Amazon Echo. With advanced microphone arrays, audio processing algorithms and ‘beamforming’ technology, devices using the NXP reference design for Amazon Alexa will be able to recognize a customer’s request from across the room, even when loud music is playing.

Beamforming, sometimes called spatial filtering, is a signal processing technique used in sensor arrays for directional signal transmission or reception. It allows IoT smart speakers to listen better, basically.

“Customers love being able to talk to Alexa from across the room with Amazon Echo,” said Priya Abani, director of Alexa. “We are excited to make it even easier for hardware makers to integrate the convenient hands-free Alexa experience into their products, while providing customers with new devices offering access to Alexa.” 

Read more: AWS makes Amazon Lex AI available to customers