X-CUBE-LocalVUI implements local voice recognition user interfaces based on audio capture, and speech recognition. It integrates Sensory TrulyHandsfree™ (THF), and Sensory TrulyNatural™ (TNL) software.
The audio capture is based on STM32 peripherals and middleware. It shows how to capture audio from the board microphone through SAI.
The example applications come with default speech recognition models. However, they can be easily updated with a user-specific model. For the examples, the specific model can be defined with the Sensory VoiceHub web tool.
X-CUBE-LocalVUI provides an implementation on an STM32H747I-DISCO Discovery kit. It can be ported to some other STM32 microcontrollers and boards.
THF, TNL, and VoiceHub are products from Sensory, an authorized STMicroelectronics partner.
- Cloudless voice user interface (UI)
- Audio capture from board microphone
- Integration of automatic speech recognition (ASR) software
- Sensory TrulyHandsfree™ (THF) software component
- Sensory TrulyNatural™ (TNL) software component
- Capability to build and integrate a customized voice UI vocabulary model
- Vocabulary model built with Sensory VoiceHub
- Support for wake word detection only, or wake word and command mode
- Detected command logged on Virtual COM port
- Possibility to connect the board as a USB device to record the microphone audio capture