On-Device TrulyNatural Speech-to-Text

Discover Sensory’s TrulyNatural Speech-to-Text (STT), the industry-leading solution that offers unparalleled accuracy in a compact size. Our embedded TrulyNatural STT engine is up to 7 times smaller than competitor engines while delivering superior accuracy. Available through the Sensory TNL SDK, TrulyNatural STT supports models for over 35 languages, with sizes ranging from 20MB to 200MB. By integrating Natural Language Understanding (NLU) models and action lookup tables, you can create a complete voice assistant in as little as 10MB. Additionally, Sensory provides the capability to incorporate domain-specific phrases, enhancing domain accuracy and minimizing model size. Experience the future of voice AI with Sensory’s embedded TrulyNatural STT—optimized for performance and efficiency on any device.

Image Alt Text

Key Features

Highly accurate with compact sizes ensuring super fast response times

Dynamically supports custom names & phrases
On-device processing ensures privacy and lower latency
Available in 35+ languages
Optimized for low power devices
Runs on mobile, automotive, and PC platforms

HOW IT WORKS

Complete Voice User Interface

Voice Assistant Capabilities: Combining Sensory’s awards winning and industry leading wake word technology, with Sensory’s domain specific Small Language Models (SLMs) Sensory can provide an embedded domain specific assistant in as little as 10MB. Or connect Sensory’s wake word and speech-to-text to a Large Language Model (LLM) running on-device or in the cloud to save money and improve privacy.

 

Image Alt Text)

Global Language Coverage

TrulyNatural STT now supports 35+ languages and regional variations including:

Afrikaans, Arabic, Belarusian. Bengali, Bulgarian, Catalan, Croatian, Czech, Danish, Dutch, English (US, UK, Kids) Finnish, French, German, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Mandarin, Norwegian, Polish, Portuguese, Romanian, Russian, Spanish, Swahili, Swedish, Thai, Turkish, Ukrainian, & Vietnamese

Image Alt Text)

TrulyNatural™: Embedded Speech-to-Text

Sensory’s TrulyNatural™ Speech-to-Text (STT) technology is designed to deliver exceptional accuracy at lower costs and can help privatize input to Large Language Models (LLMs). TrulyNatural STT addresses these needs by optimizing its performance across multiple device resource variables, including memory, compute speed, and power, to provide a high-quality voice user experience for any edge platform.

Request the TrulyNatural™ STT product brief to learn more.

Image Alt Text)

Stay Connected

Get the latest speech recognition and artificial intelligence resources and industry updates.