Performing speech recognition at the edge, rather than sending data back to the cloud, is a major engineering challenge. You need significant processing power on a tiny energy budget, and often in a small form-factor. In this episode of Chalk Talk, Amelia Dalton chats with Gerard Andrews of Cadence Design Systems about the Tensilica HiFi 5 DSP core, which you can drop into your SoC design to give you powerful speech recognition capabilities.
Click here for more information about Tensilica HiFi DSPs for Audio, Voice, and Speech