AVC202 series speech recognition chip

Product overview

The AVC202 series, as a low-power single-microphone offline speech recognition SOC chip, has the ability to support all the way to analog microphone audio input. The chip integrates key components such as 32bit RISC processor, FPU and audio ADC, and is also equipped with other rich peripherals. It is worth mentioning that the chip specifically joined the DSP instruction set, which can significantly improve the speed of audio signal processing and offline speech recognition processing.

AVC202 series chips have built-in offline speech recognition algorithm, which can be quickly deployed and applied without complicated development. It can be widely adapted to various products such as voice remote controller, intelligent wall switch, smart watch, smart headset and smart glasses, helping the equipment to quickly realize voice interaction function and bring convenient intelligent experience to users.

Application areas:

Voice recognition remote control

Voice switch

Smart Watch

Smart headphones

Smart Glasses

Product function introduction

Kernel

High-performance 32-bit RISC core

Support DSP instructions, support FPU floating-point arithmetic

Storage

Built-in SRAM

Built-in 16Mbit FASH

Audio

Supports 1 analog microphone input or 2 digital microphone inputs

Clock and Reset

Built-in 24MHz RC vibrator

Built-in 32KHZ RC vibrator

Support watchdog

Support power-on automatic reset

Peripherals

One SPI Master/Save interface

2 UART

2 I2C master/slave controllers

Power

Power Input Range 1.8V ~ 3.3V

Built-in DO

Working temperature

Ambient working temperature:-40 ℃ to 85 ℃

SDK

Provides a complete library of end-side speech recognition algorithms.

Provide UART communication protocol

Encapsulation

QFN20-3x3mm