Snips partners with DSP Group for multi-room edge voice recognition architecture
Wireless chipset company DSP Group and embedded voice recognition software provider Snips have announced a technology partnership to combine Snips’ biometric technology with Ultra Low Energy (ULE) system-on-chip (SoC) products and modules. The collaboration is intended to extend the reach of natural voice-controlled edge devices, and allow OEMs to leverage a central hub to place satellite node devices with speech recognition and streaming capabilities in any room of a home or office.
Snips launched its enterprise product suite earlier this year, which consists of Flow for voice interfaces, Commands for embedded voice commands, Wake for wake words, and Satellite for voice streaming to Snips Flow-enabled devices. The company says it is the first to provide solutions ranging from voice commands on microcontroller-based hardware to natural spoken language understanding (NLU) running on microprocessor-based hardware.
“Snips and DSP Group share a vision for the future of embedded voice recognition across multiple verticals. Edge technologies play an important role as the adoption of voice recognition technologies increase. By combining our voice recognition software, running on-device and offline, together we give OEMs a new possibility to create a local grid of mutually-aware devices, lowering the BoM per device and giving opportunities to diversify their product offerings,” said Yann Lechelle, Chief Operations Officer at Snips. “This new voice architecture will be applicable in smart home and smart building verticals where performance, cost effectiveness and data privacy play an important role for end consumers.”
DSP Group’s ULE/DECT technology provides secure network connections up to 500 meters, and can overcome interference, such as in noisy environments, to make it a robust technology for voice transmission compared to the 2.4 GHz frequency band, which can become noisy with WiFi and Bluetooth communications, according to the announcement. Snips technology can also run on an offline device, keeping personal data from being transmitted over the internet.
The global market for voice and speech recognition is predicted to rise from $1.1 billion in 2017 to $6.9 billion in 2025 by Tractica.