Font Size: a A A

The Design Of Customized Ringing Tone Customization Platform Based On Speech Recognition Technology

Posted on:2017-03-30Degree:MasterType:Thesis
Country:ChinaCandidate:Y L GuoFull Text:PDF
GTID:2348330518472344Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
Speech recognition technology is an important field of the development of computer technology, is a machine can realize the voice signal is automatically converted into text signal. In other words, the machine can understand human language, and in accordance with the wishes of the people work. Speech recognition technology needs to be combined with linguistic, acoustic, digital signal processing, artificial intelligence and pattern recognition,and other aspects of knowledge.Firstly, the basic knowledge of this thesis, a brief introduction of CRBT service,describes the timing, modify, cancel CRBT process, appear to have the limitations of mass customization.Secondly, from the basic model of speech recognition technology, this paper introduces the principle of pattern recognition and digital signal processing in the aspects of endpoint detection, pretreatment, module generation, module matching and module management. The HMM (hidden Markov model) and DTW (dynamic time warping) speech recognition system design of the Chinese language. HMM has a strong ability to time sequence modeling, the characteristic parameters of the training, each speech corresponding to their hidden Markov model, speech and hidden Markov model corresponding matching, to complete the identification process. HMM is often used in the need to identify a large vocabulary situation.DTW can deal with voice signal characteristic parameters of time length is not isotropic problems, with fast recognition speed, small system cost less and effective vocabulary processing. At the end of each section are simulated by Matlab.The final design of the platform customized ringing tone based on speech recognition technology. Customization platform by using acoustic CRBT system IVR architecture design technology, combined with the key technologies in the field of speech recognition and speech synthesis, speech, using voice XML standard language tools for the development of an application system. The user by entering a specific number into the server, use speech or button selected singer or song in the form of transfer to the system command, you can get their own customized color ring back tone. Through the use of a dedicated language interface to facilitate the user's operating process, the main use of VXML technology to achieve the phone process.
Keywords/Search Tags:speech recognition, feature extraction, hidden Markov model, dynamic time warping, customized ringing tone speech customization platform
PDF Full Text Request
Related items