Voice identification circuit and translation system

Abstract

The invention discloses a voice identification circuit and a translation system. The voice identification circuit comprises a circuit board, a microphone array, a voice identification module, a microprocessor module, an audio frequency operational amplifier module, a loudspeaker and a voltage dividing module. The microphone array, the voice identification module, the microprocessor module, the audio frequency operational amplifier module and the loudspeaker are arranged on the circuit board and are sequentially connected with one another, the voltage dividing module is connected with the voiceidentification module, a plurality of control input ends are arranged on the voice identification module, the voltage dividing module comprises a plurality of voltage dividing branches, the voltage dividing branches are arranged between a power positive pole and a power ground in a connected manner, are connected with one another in parallel and comprise a plurality of resistors, the resistors are serially connected with one another to form the voltage dividing branches, and the various control input ends are connected with the various voltage dividing branches. The voice identification circuit and the translation system have the advantages that the voice identification circuit is low in cost, and the voice identification ranges and noise reduction effects can be dynamically balanced.

Description

technical field [0001] The invention relates to the technical field of speech recognition, in particular to a speech recognition circuit and a translation system. Background technique [0002] At present, the working process of translators on the market is as follows: First, a dual-mic array, four-mic array or eight-mic array is used to pick up the sound source, form a voice stream, and perform preprocessing, pre-filter the voice stream, and retain 300 -3400Hz voice signal, and then perform A / D conversion and anti-aliasing processing to convert it into 8bit, 16bit, 32bit digital voice signal, and then pass DSP processing to complete DA conversion, and finally amplify through audio amplifier to output audio signal. [0003] However, the voice recognition systems in translators currently on the market have problems such as echoes and misrecognition. The reason for the above problems is that there is interference in the voice recognition system. The interference usually include...

AI Technical Summary

Problems solved by technology

[0003] However, the voice recognition systems in translators currently on the market have problems such as echoes and misrecognition. The reason for the above problems is that there is interference in the voice recognition system. The interference usually includes the following parts: (1) The interference of the circuit itself: Since the audio signal input terminal belongs to the weak signal of the microphone, it is easily affected by interference sources such as strong signals around; (2) Interference of power supply: In the translator, there are several power supplies of 5V, 2.8V and 1.8V, high voltage and Low-voltage circuits are easy to interfere with each other, and will couple the interference to the audio transmission signal; (3) Ground interference: Since there are audio signals and high-frequency signals in the translator, it is easy to cause the ground signal to bounce, and the ground signal returns to the process will interfere with the audio transmission signal

Images