Font Size: a A A

Design And Implementation Of Smart Air Conditioner Speech Recognition System

Posted on:2021-02-26Degree:MasterType:Thesis
Country:ChinaCandidate:B TanFull Text:PDF
GTID:2392330614971488Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Dialogue is the most important and efficient way of communication in nature,especially for human beings.The vast majority of human social behavior is conducted through voice.At the same time,with the rapid development of artificial intelligence and other technologies,voice has begun to be transplanted to other fields other than humanto-human communication,including software and hardware in the Internet industry.The continuously mature speech recognition technology not only makes the communication between people and machines possible,but also smoother and more efficient.The rapid development of the Internet of Things technology has also made Internet technology widely used in the traditional mechanical field,such as the combination of voice recognition technology and air-conditioning electronic control system,changing the traditional air-conditioning control mode to achieve real human-computer interaction.This paper studies the design and implementation of the smart air-conditioner speech recognition system,the main work is as follows:The author participated in the system demand analysis in the project design of the smart air-conditioner speech recognition system,and analyzed the functional and nonfunctional requirements of the smart air-conditioner speech recognition system through investigation,and designed the system according to the needs.The system is mainly divided into a voice processing module,a voice recognition module,an air conditioning control module and an intelligent recommendation module,and the functional requirements of each module are analyzed.Afterwards,the main functional modules of the system were designed,including the speech processing module's collection of speech,and the realization of AEC(Automatic Echo Cancellation)based on Web RTC,and endpoint detection.The feature extraction in the speech recognition module is based on Kaldi to build ASR(Automatic Speech Recognition,automatic speech recognition technology)for acoustic model training and speech decoding.Finally,the system is combined with an air conditioner for functional testing to ensure the stability and reliability of the system.This paper mainly studies how air-conditioning companies use existing mature speech recognition technology to design and implement a speech recognition system suitable for intelligent air-conditioning,rationally use voice recognition technology to serve traditional household appliances,and build a software and hardware implementation plan.At present,the system has been tested in combination with the whole machine,and it is preliminarily determined that it has certain market competitiveness,which can bring economic benefits for the subsequent company's smart air conditioning project.
Keywords/Search Tags:Speech Recognition, Automatic Echo Cancellation, Kaldi, ASR
PDF Full Text Request
Related items