VSR-02 embedded speech/speaker recognition module

1.  Product description:
VSR-02 embedded speech/speaker recognition module is designed for a wide variety of applications with speech/speaker recognition function. The training mode allows add speech commands/passwords to user-defined vocabulary. After the training when you speak to this module a word that has been trained, this module will take a defined action associated with the spoken. Equipping with this module allows control a device, appliance or toy just by speech. The module uses the advanced technique for electronic design and recognition algorithms, so that has high accuracy, short latency and good noise-robust performance.

2.  Specifications:
     a)   Speech commands can be customized by training mode in any language
     b)   Speech vocabulary size up to 64 (default 10)
     c)   Recognition accuracy rate 98%
     d)   Response latency approximately time 0.5
     e)   Maximum duration of speech commands is 2.5 sec
     f)   Power supply 3.5 - 5.0 VDC
     g)   Operating temperature 0-70 degree
     h)   Dimensions 40 X 67 X 7.5 (mm)

3.  Interfacings
  1) Training interface
VSR-02 embedded module provides an UART (comport) that is used for training interface, parameter settings and data communication with any host computer.
  2) Driving interface
This embedded module provides a connector that has GPIO (24 channels), PWM (4 channels), I2C and SPI; all these can be interfaced with external circuit of driving/control for the filed applications.

4.  Applications:
VSR-02 embedded module has a wide variety of applications for speech recognition, it includes:
     a)   Speech controlled toys and appliances
     b)   Speech controlled entry system
     c)   Speech assisted computer games
     d)   Speech assisted disabilities
     e)   Speech assisted driving