Audio specialist Knowles Corp. has announced the launch of a development kit aimed to ease voice integration for the Internet of Things — and it centers around a Raspberry Pi single-board computer.
"Knowles designed this new kit to be the simplest and fastest way for product designers to prototype new innovations to address emerging use cases including contextually aware voice, ML [Machine Learning] listening, and real-time audio processing, that require flexible development tools to accelerate the design process, minimize development costs, and leverage new technological advances," claims Vikram Shrivastava, senior director for IoT marketing at Knowles, of the launch.
"By selecting Raspberry Pi as the system host, we are opening up the ability to add voice and ML to the largest community of system developers that prefer a Linux or Android environment."
The AISonic development kit is built around Knowles' AISonic IA8201 Audio Edge Processor, which features two Tensilica digital signal processing cores — the first designed for performance-focused workloads including artificial intelligence at the edge, and the second designed to draw as little power as possible for always-on sensor needs.
Two microphone array boards are bundled with the kit, using Knowles' Everest microphones, while software support is provided through the OpenDSP platform with libraries provided for voice-wake, beamforming, custom keyword detection, and noise elimination, among others. TensorFlow Lite support is also provided.
Physically, the kit is split into three boards — each connected via ribbon cable. a HAT board connects to the Raspberry Pi's general-purpose input/output (GPIO) header and breaks off to a processor board holding the IA8201 chip, which in turn connects to one of the two bundled two-mic or three-mic microphone array boards.
Pricing for the kit was not published at the time of writing, with more information available on the Knowles website.