In ipXchange’s second interview with AONDevices at CES 2024 – check out the first one here – Guy chats with Connor for a live demonstration of this ultra-low-power, AI-enabled sound recognition technology.
As Connor shows us, AONDevices’ evaluation kit can be used to create a private voice assistant that recognises several wake-up voice commands (keywords or phrases), thanks to the on-chip AI algorithm running at <260 µW. Once a wake-up command has been detected, AONDevices’ chip can then use this as a trigger to control aspects of a larger system, such as a smart home device.
Additionally, AONDevices’ solution can detect and classify continuous sound sources, in this case music. This functionality extends to recognise sounds such as crowd noise, traffic noise, baby or animal noise, or mechanical noise, and makes sure that voice commands are still detected – and work – when consumers use your product in these types of environments.
Try this technology out for yourself by following the link to the dedicated board page and filling out a form with details of your commercial project, or tick the NDA box if you’re working on something a little more hush-hush.