‘Twas the night before Christmas, when all through the house
Not a creature was stirring, not even a mouse;

OK but are you sure there was not even a mouse stirring? Maybe you need some better hearing technology in order to pinpoint the exact location of said mouse. In which case, you could certainly whip-up a demo that can identify the location of a stirring creature anywhere in the room.

That’s the EYE on NPI we have for you this week, an Acusis S linear microphone array with specialized support hardware to let your voice recognition software worry less about echo cancellation and more to the code you want to write!

With the ease of adding voice-assistant technology to your products, thanks to powerful back-end voice recognition engines and improvements in edge voice recognition, you can add voice control without needing a bunch of PhD’s in machine learning. The toughest problem is how to get clean audio to your engine. Unlike video recognition, which sometimes just needs good lighting added, you can’t ‘add’ anything to audio to clean it up. Being able to capture audio as well as possible is going to make a huge difference in whether your voice system works wonderfully or just ‘so-so’, especially in the final installation which can have background noises like traffic, TVs, pets, or children.

The thin PCB is plug-and-play with a USB. We plugged it into our MacBook and it showed up instantly as a microphone input. Our Windows computer didn’t seem to like the composite device but it could be we have a odd Windows 10 install with all our libusb drivers from years of hardware hacking. On the board are 4 microphones and an XMOS DSP chip that runs VocalFusion XVF3500 far-field voice capture and DSP processing. There’s a row of LEDs that help you determine where audio is coming from which is a nice little add-on. There’s mounting holes all over the place, and a rear-mount USB connector for flat panel-mounting.

Antimatter Research’s Acusis S linear microphone array includes PDM microphones, XMOS XVF3500 DSP, and Philips BeClear™, working together to enhance far-field audio input for crystal-clear voice recognition. A computer will see Acusis S as a microphone as well as a speaker, allowing users to optionally pass their computer audio through it via the audio jack for echo-cancellation. It also comes with a MacOS™/Linux™/Windows™ configuration tool that allows users to fine-tune Acusis S for their desired application and environment. Acusis S works out-of-the-box with a USB cable.

Benefits and Features

• Far-field reception
• Automatic echo cancellation
• Voice control
• Beamforming
• Dereverberation
• Noise suppression
• Philips BeClear Speech Processing
• Stereo or mono pass-through
• MIC audio out
• Mute with indication
• Plug and play
• Tunable
• USB, UAC1, and UAC2
• PDM microphone array with XMOS DSP

You can pick up the Acusis S for instant gratification and integration into your voice recognition system – since it just shows up as a USB audio device, you can use it with any software on any OS. Digi-Key has plenty of Acusis boards in stock and can ship same day so you can start your prototyping tomorrow morning!

See this at Digi-Key short URL https://www.digikey.com/short/z112nf

Catch Antimatter Research on Twitter: https://twitter.com/antimatterai

See the three Antimatter Research videos below: Video 1, Video 2, Video 3

https://www.youtube.com/watch?v=NbiGoK2romY

https://www.youtube.com/watch?v=H8CA_K8l3lc