Authorization Announces Embedded Single Engine Multi-WakeWord and Speech Commands Recognition Solution for Micro-CPU Architectures

MONTREAL, Feb. 22, 2019 /PRNewswire/ --A, a speech recognition provider offering accurate, fully embeddable speech recognition solutions in any language, announced today the release of its ultra low-power voice control solution for embedded platforms such as Cortex M4F processors. This breakthrough new solution delivers, using a single engine, multiple wakewords in parallel and up to 50 multilingual commands with variable phrasing on a footprint equivalent to 50 MHz CPU, 128KB of RAM, and 550 KB of flash storage - a quantum leap over currently available software's newest small footprint solutions enable the proliferation of voice user interfaces in previously infeasible scenarios such as smart light switches capable of recognizing offline speech commands. Global OEMs and ODMs in the smart home, smart toys, mobile devices, IoT, manufacturing and other industries can now enhance their own offerings and access new markets quickly with's embedded, multilingual speech understanding systems. A video demo of's embedded solution can be accessed at's partners are also extremely excited about this breakthrough. David Huang, A SVP of TCL TONLY Electronics Holdings Limited, a leading global ODM of audio-visual and smart home products, said, "This is a game-changing innovation for those requiring low latency, low footprint, offline voice command solutions. We are happy to partner with to bring the next generation of products to our joint customers built on top of's flexible software platform. This will accelerate the application of voice in smart home and IoT markets around the world." will be at Mobile World Congress next week in Barcelona to showcase their next generation of on-device speech understanding systems.To schedule a meeting, please email develops highly accurate and intuitive speech understanding solutions in a small footprint and low-latency package capable of running offline on small devices.'s unique and patented speech-to-intent approach provides faster, more accurate and noise-robust speech recognition in any language.'s solutions enable consumer electronic device manufacturers and OEMs to develop unique and differentiated voice user interface solutions for their devices. Announces Embedded Single Engine Multi-WakeWord and Speech Commands Recognition Solution for Micro-CPU Architectures
View original content:
See also:
Leave a comment
  • Latest
  • Read
  • Commented
Calendar Content
«    Май 2021    »