Core Building Blocks
A first of its kind AI-solution that uses unsupervised learning to correlate audio data and get smarter over time. Our unique design approach does not use modeling or extensive training data but rather uses inferencing AI to adapt to the dynamically changing auditory scene.
A voice-biometric driven tracking methodology for effective voice from noise and voice from other voices isolation. A byproduct of this approach is the ability to extract perceptual data (identity, emotion, direction, range, gender) from an individual voice within a noisy signal.
Signal Enhancement IP
IP that repairs and enhances the signal of interest for a better listening experience, improved performance for automatic speech recognition (ASR), effective metadata extraction, and accurate speaker identification. In addition, our unique method of signal enhancement does not introduce unnatural artifacts.
The Yonder Product platform adds value to the entire voice ecosystem fundamentally changing the effectiveness of edge based VOICE technologies.
Yonder CAFE (Conversational Audio Front End)
Platform agnostic (Google, Amazon, Watson, Etc) edge-based for enhanced ASR/NLU capabilities in high-noise environments.
Adaptive Biometric Identification
On the edge “Voice Print” generation (voice biometric identification) capabilities for user authorization and profile retrieval in high noise environments.
Adaptive Wind Noise Solution
Yonder’s CAFE with Wind Noise intelligently adapts its signal processing when the solution detects wind interference.
Adaptive Discriminant Listening
ADL discriminantly distinguishes between simultaneous signal contributions of multiple audio sources and produces an artifact-free, authentically reconstituted signal where the desired voice is pushed to the auditory foreground while the remaining source contributions are relegated to the auditory background.
User Barge-In Solution
UBI allows voice command users to interrupt or barge-in while a device is in media play-back mode (speakers engaged) or in situations when the device is generating its own noise interference (e.g. household appliances, drones, surveillance cameras).
Adaptive Command Recognition
Limited On-the-edge Speech to Text capabilities for effective keyword and command recognition that provides hands free device navigation in high noise environments.
Adaptive Directional Listening
This solution was designed for effective far-field voice capture in dynamically changing noisy environments. ADL leverages our voice DNA tracking abilities with user-defined directionality. As a platform customizable solution, it is well suited for defense, surveillance, and security-focused platforms.
A Convergence of Yonder Capabilities
VISPR (Voice Isolation for Sonic Perceptual Recognition) is a first of its kind auditory processing system that adds a layer of “perceptual awareness” to voice platforms.