Offline Voice Control for ARM and X86 Platforms
vicCONTROL industrial is the most powerful member within our vicCONTROL voice control family. It's available for ARM- and X86-based platforms under Linux, Windows, Android and iOS.. Due to the increased computing power of embedded processors and their widespread use in device and machine control, vicCONTROL industrial can be used directly without additional processor hardware.
vicCONTROL industrial works without an Internet connection, since the voice control is implemented locally on the processor. It is therefore particularly suitable for applications that must function without Internet access or have special data protection requirements (privacy by design).
vicCONTROL industrial recognizes natural speech, comparable to cloud-based speech recognizers such as Amazon's Alexa or Google Speech-to-Text Services. Additionally it is available in 30 national languages and enables semantic understanding of what is being said using the latest machine-learning / artificial intelligence technologies.
Development kits are available forARM- and x86-compatible platforms. In addition, you get access to our easy-to-use web tool vicSDC, which allows you to create your personal speech dialogs.
An interface based on the IoT or web standards MQTT / JSON is available for communication between voice control and your application.
- Voice Control software for ARM- / x86-compatible platforms under Linux, Windows, Android and iOS
- Industry-proven quality in 30 national languages
- NLU-capable voice control
- Creation of statistical and semantical models for offline operation via state-of-the-art ML and AI technologies
- Convenient connection to applications using MQTT and JSON APIs
- Semantically evaluated commands are passed on in a structured manner
- Freely selectable wake words
- Development kits for ARMv7 and x86-compatible platforms available
- Delivery of a development kit includes an access to our web tool vicSDC for comfortable and easy speech dialog creation
- All voice commands are available at the same time
- Extraction of multiple semantic parameters from a single command, regardless of their sequence
- Qualitative evaluation and prioritisation of recognition results
- Acoustic feedback through freely definable audio prompts
- Multiple activation modes by voice or external trigger
- Extension of GUIs with voice control
- Control of machines and appliances
- Operating and monitoring of medical systems
- Intuitive input in rehabilitation technology
- Natural collaboration with robots and in AR applications
- Self-explanatory user interfaces
- Runtime: from 6 MByte
- Speech model: max. 6.3 MByte per language
- Application-specific language resources:
- approx. 10 kByte simple command control with 10 words
- approx. 500 kByte extensive control with 1000 words
- RAM: from 16 MByte (depending on application)
- Computing requirements: from 400 MIPS fo real-time processing