We have launched a Bluetooth voice recognition module with high cost performance. It is highly integrated with local voice system chip (SoC), and self-developed local wake-up and front-end noise reduction algorithm based on neural network deep learning. No need for additional license.

  Multiple level wake-up. Multiple local fixed word command words. Supports multi-channel / multi mic echo cancellation function. It can be used as AEC front-end processing + local wake-up. Built in deep customized neural network processing unit and logic control unit. No need for MCU to handle simple logic.

• Offline identification, no need for networking
• Self-defined voice entry settings
• High voice recognition rate, low false trigger rate