ClearVox is a software suite of advanced voice input processing algorithms aimed to enhance voice clarity in any voice-enabled device.
Voice has become the primary user interface for services such as cloud-based conversational assistants and voice activated IoT, and it requires a clean voice signal in any given environment to ensure accuracy and intelligibility by smart speakers, Bluetooth speakers, headsets, mobile phones, wearables, infotainment and smart home systems.
ClearVox is leveraging Cevas vast experience in audio and voice processing and incorporates advanced algorithms that cope with different acoustic scenarios and microphone configurations, including optimized software for speaker direction of arrival, multi-mic beamforming, noise suppression, and acoustic echo cancellation.
Front-end voice processing software package providing enhanced speech intelligibility for voice-enabled devices
Overview
Key Features
- ClearVox incorporates multiple algorithms such as SW VAD (Voice Activity Detection), direction of arrival, beamforming, noise suppression and acoustic echo cancellation
- ClearVox is offered in two pre-packaged software form factors
- Near-field applications including headsets and wearables
- Far-field applications including smart speakers, conference rooms, infotainment and other advanced voice-enabled applications
- The accompanying SDK is available for customers to configure and adopt ClearVox to their unique product specifications, such as number and geometry of microphones, echo tail length and more.
- ClearVox is available exclusively for Ceva's sound DSPs: Ceva-TeakLite-4 (ultra-low-power) and Ceva-X2 (high-performance)
Benefits
- Software suite of advanced voice input algorithms for Ceva sound dsps, targeting smart voice activated devices. Significantly reduces barrier-to-entry for voice-enabled device makers
- Configurable and modular to allow customization for silicon vendor and OEM to address specific product needs such as number of mics, mics geometry and echo tail
- Unique software architecture to minimize computation workload for efficient cost & power
Block Diagram
Video
MWC BARCELONA 2018 CEVA CLEARVOX DEMO
Far Field Voice Controlled smart speaker reference design
Applications
- Smart speakers
- Voice Controlled IoT
- Bluetooth speakers
- Headsets
- Mobile phones
- Wearables
- Infotainment
Technical Specifications
Maturity
In Production
Availability
Available
Related IPs
- Neural network-based speech recognition technology for voice assistants and IoT devices
- Enhanced Neural Processing Unit for safety providing 32,768 MACs/cycle of performance for AI applications
- Enhanced Neural Processing Unit for functional safety providing 49,152 MACs/cycle of performance for AI applications
- Low-power, low-gate-count, highly-configurable DSP core for audio and control processing
- Advanced communication DSP, providing cutting-edge performance for multi-Gigabit class connectivity
- Tamper-resistant Cortex-M processor with optional software isolation using TrustZone for Armv8-M