Voice Control

Voice control has become indispensable in many areas of everyday life. The clear advantages of voice input and output for accelerated, intuitive and contactless device operation also apply to professional applications. HY-LINE offers individual solutions for your professional and industrial applications.

While the output of signals and sounds is supplemented by synthesized speech, voice control takes its place during input. This allows devices to be operated conveniently without hands and without having to look at the device, turning the machine into a voice assistant.

Voice control is used in particular where hands are not free to operate the device, are dirty or must be sterile and must not touch anything. Offline voice control does not require an internet connection during operation. This means that data remains private and secure and can be entered in real time. The data does not have to be loaded on an external server, nor does permanent broadband access to the Internet have to be provided.

Technical requirements

Voice control runs on ARMv7 and higher as well as x86 platforms under Linux, Windows, Android and iOS. The voice control can run on the system processor. Typical system requirements are

• Main memory: from 16 MByte (depending on the application)
• Computing requirements: from 400 MIPS for real-time processing
• Base Runtime "Dialog Manager": 10 MB
• Speech recognition, NLU and semantic models together per language: 10 MB
• Text to speech synthesis per language: 100 to 200 MBsorted protocols (MQTT, JSON, (S)RTP) with your device application

Implementation

• Design your individual speech dialog with the help of the provided WebSDK (Web-based Software Development Kit)
• Deploy the voice application created in this way on your embedded system. We will be happy to advise you on suitable solutions
• Connect the voice application to your device application via standardized protocols (MQTT, JSON, (S)RTP)

To meet the requirements for a professional system and at the same time to make the effort for the implementation of your individual voice application directly on your local device as simple as possible, HY-LINE has the Voice Control StarterKit developed, which makes the way to a professional voice control much easier and faster, with the development directly on an industrial-grade embedded platform of the latest generation incl. High-end touch display.

The starter kit offers:

• Industry Grade: pITX-iMX8M-Dual-Core Single Board Computer (including Passsivkühlung) with high reliability and availability
• High End Touch Display: HY-LINE 7" TFT Family Concept Display including PCAP-Touch
• Voice Control Software: Industry-proven quality in 30 languages and with wake-word function
• Text to Speech: now with offline speech synthesis for output of any text
• Plug & Play: Linux Image pre-installed on Micro SD Card including bracket for easy handling
• Demo App: with pre-installed sample applications for voice control and synthesis
• WebSDK: Webtool dialog designer for the implementation of natural language communication without programming