The Apollo4 Audio Kit is a flexible hardware platform designed to support the Apollo4 Family Voice-on-SPOT® (VoS®) Kit.
VoS enables professional AI endpoints with automatic speech recognition. It is based on Ambiq’s Apollo4 Plus and Blue Plus SoCs which deliver power efficient MCU functionality and ample memory to enable quality voice and audio on battery operated embedded systems with a tiny power budget.
AI application developers can use the AmbiqVoS SDK and Ambiq’s proprietary voice activity detection (VAD) algorithm to detect human speech and use their preferred audio preprocessing solutions for noise reduction and beamforming. Users can harness the potential of VoS’s high-compute, low power capability with highly optimized keyword spotting models for accurate voice command and control for local and cloud-based applications.
VoS Feature Highlights
• One stereo analog microphone and four stereo digital microphones for truly always on listening
• Includes Ambiq's propriety VAD algorithm and supports Sensory's TrulyHandsfree wake word engine and DSP Concepts beamforming algorithm
• Supports Alexa Mobile Accessory (AMA) protocol and Android TV Voice Services
• Passes Alexa Voice Services (AVS) Far-Field and Near-Field testing for FAR, FRR, RAR
• Audio Compression: Opus: mSBC: ADPCM
• Third Party Audio Processing
- DSP Concepts 2-MIC beamforming SCNR audio preprocessing
- Sensory TrulyHandsfree® (THF) MCU wake word engine
The AMA4AUD Audio Kit Contents
• Apollo4 audio shield
• Invensense analog microphone
• Vesper digital microphone
• Memsensing digital microphone
(Note: Apollo4 SoC evaluation boards are not included)