Pegatron Adopts XMOS's Voice Technology
Pegatron has incorporated XMOS's VocalFusion voice technology into its smart voice assistant Martina.
Pegatron's smart voice assistant Martina, part of its AI portfolio of solutions, uses Google Cloud Services and natural language recognition technology to support Chinese, English, Japanese, and Korean. The XMOS XVF3000 voice processor deployed in the Martina unit uses acoustic echo cancellation (AEC) to remove playback audio from the microphone signal, allowing users to barge-in and talk over the playback signal. Noise suppression and dereverberation technologies remove unwanted noise and echoes in the captured signal, providing a crystal clear output signal. The microphone follows a voice as it moves around the room (via beamforming), and ensures that even the quietest person in the room can be heard.
"The launch of this device further extends the deployment success of the XVF3000 voice processor and is testament to the XMOS comprehensive far-field voice interface portfolio," said Mark Lippett, president and CEO of XMOS, in a statement. "Voice-enabled smart speakers, like Pegatron's Martina, are set to see significant growth, and we are proud to have worked with Pegatron Corporation on this project."
"The XMOS solution was selected for use in our smart voice assistant due to its superior far-field voice technology," said Joe Wu, senior director at Pegatron, in a statement. "Martina, with its cross-platform compatibility and voice-driven user interface, has the ability to connect and control a wide range of devices and services using simple voice commands. The quality of the voice processor is therefore critical to its success as a product, and the XMOS solution truly meets our performance requirements."