Xmos has introduced a new developer kit for Amazon Voice Service (AVS) that includes a far-field two-mic solution optimized for smart TVs and set-top boxes.
AVSs are devices that have a microphone and a speaker that users can talk to directly with the wake word Alexa and receive voice responses. These devices feature built-in products that work with Alexa skills and Alexa-compatible smart home devices.
The dev kit is based on Xmos’ XVF3510 voice processor built for modern living spaces. The technology identifies and isolates a voice command from other sound in the room including media streaming through the device, enabling far-field voice capture with close range precision.
The algorithms include:
- A stereo acoustic echo canceller that suppresses unwanted speaker echo and enables barge-in.
- Interference canceller that uses null point noise sources to cancel out unwanted background noise.
- An adaptive delay estimator that adjusts audio signal latency to ensure the acoustic echo cancellation algorithms deliver a real-time experience.