EveryWord ™ Audio Front End (AFE) Voice Processing Module
ULTRA FAR-FIELD DSP AUDIO FRONT END (AFE) FEATURING
CIRRUS LOGIC SOUNDCLEAR® and SENSORY VOICE INTERACTION TECHNOLOGY
Audio Front End (AFE) Features
- Cirrus Logic’s SoundClear voice technology
- 7 to 9 mic performance using only 3 to 4 mics
- Exceptional capture range of (>9 meters)
- H/W accelerated royalty-free Amazon wake-word engine
- Compatible with most PDM mic arrays
- Flexible mic array spacing and orientation (2D, 3D, vertical, horizontal)
- Onboard clocking
- For applications leveraging existing host/computer board
- Ark X host F/W for integration
- Eliminates need for high density PCB
- Amazon AVS qualified (both 2 and 4 mics)
EveryWord ™ AFE MODULE
The AFE Module can transform any existing product into a voice enabled product.
Clear Performance Advantages
- Captures voice commands:
- from across the room (>9 meters)
- in noisy and reverberative environments
- during loud music/movies
- Unique ability to discriminate and suppress speech from TV
- Flexible placement: wall, in-wall, ceiling, dashboard
- Highly-integrated production-ready module
- Works with most existing mic array designs
- Amazon AVS qualified (both 2 and 4 mics)
- Ultra-low power battery operation for wake-on-word
Technical Overview
- Supports linear, Y, square, triangular and 3D mic array geometries
- Proprietary 3D audio processing technology significantly outperforms beam-forming solutions
- Full-duplex far-field voice communication
- Does not require source ducking for reliable interaction
- Concurrent voice service and voice communication
- Multiple user tracking
- 4x bidirectional I2S/TDM audio ports
- 2x host to AFE
- 2x available for external connection
- 12 AECs for superior barge-in and multichannel audio
- AEC support from mono to 5.1 channel surround
- GPIO, SPI and I2C connectivity
- Web server-based console
EveryWord ™ AFE Module System Overview
AFE Quick Development Solution for Existing Platforms
