TalkTo™ is a suite of proprietary Machine Learning & Microphone Processing algorithms that delivers a quiet-room experience in the noisy & unpredictable spaces of everyday life.
To support the full spectrum of voice-enabled products, TalkTo is available in multiple variants, each scaled and optimized for the needs and capabilities of the target application.
Set Top Box
Hearable / Wearable
TV / Smart Screen
Smart Home / IoT
Anatomy of a Voice-Controlled Product
Audio Front End (AFE)
Conceptually, the AFE serves as a microphone cleaner. It uses the raw, noisy audio from the microphones to detect, extract, and clean any speech-activity found in the ambient sound field.
Wake Word Engine (WWE)
Wake-Word Engines continuously scan the AFE-provided voice-stream, looking for the presence of a specific phrase (e.g. "Alexa"). Accuracy of the WWE is a primary driver of UX-quality and depends heavily on AFE quality.
Back End Services
After the wake-word is detected, the extracted voice-command is passed to a [cloud-based or local] voice-service which converts speech-to-text (ASR) and determines user-intent (NLU). Again, accuracy is highly dependent on AFE performance.
For robust Voice UI performance, products with audio-output(s) must have low speaker-distortion. To achieve this, output-audio is digitally limited to the physical capabilities of the speaker(s). Audio Weaver® is a fast path solution to this problem.
Thanks to DSP Concepts' exclusive Adaptive Interference Canceller™ (AIC), TalkTo™ is able to detect and extract faint voice-commands, even in noise-saturated environments.