Voice UI Features
Trigger/wake word recognition
Audio Weaver® voice UI modules can perform recognition of wake word(s) the product designer chooses, and can be easily tuned to achieve the most reliable wake word recognition from a given set of components.
Direction of arrival determination
Through precise signal gating and automatic optimization to get the best possible performance with a given ambient noise level, our voice UI processing can accurately determine the direction of a user’s voice—and ignore room reflections and stray noises.
Our acoustic echo canceller not only cancels the sounds coming directly from the speakers built into a voice command product, it also cancels the reflections of those sounds from floors, walls, ceilings and objects in a room.
Voice UI powered by Audio Weaver® uses multiple microphones to focus on sounds coming from the user’s direction and ignore other sounds. The beamformer design is based our extensive research into the characteristics of the many types of mic arrays.
Our algorithms identify and cancel various types of environment noise. The noise reduction function can be used at multiple points in the processing chain, such as in front of the wake word detection algorithm and after the beamformer.