r51due

英国essay论文精选范文：“浅谈语音识别技术的发展”，这篇论文讨论了语音识别技术的发展。近20年来，随着信息技术的不断发展，语音识别技术也逐步发展成熟。语音识别技术可以让人更加方便地享受到更多的社会信息资源和现代化服务，对任何事都能够通过语音交互的方式。近20年来，语音识别技术取得了显着的进步，并逐步的走向市场。在未来的日子里，语音识别技术将应用更为广泛。

50, most researchers are committed to the basic concepts of acoustics and phonetics. In 1952, in at& t Bell Laboratories led to a single pronunciation of isolating the pronunciation of 10 English digit speech recognition system, method is to measure the Yuan Yinyin formants of each number. In 1956, RCA lab conducted a separate study, 10 different syllables to attempt to identify a single pronunciation, uses the same method of measuring formant. In 1959, the United Kingdom University College researchers begin to try another way, using spectrum analysis method for phoneme recognition and template matching to identify 4 9 consonants and vowels. 60, appeared and published several important ideas for speech recognition. In 1962, the University of Tokyo was another phoneme recognizer hardware implementation attempts. Using zero-crossing rate method of separating different parts of the speech signal, and recognition. In 1963, Japan NEC lab for digital voice recognition attempt can be regarded as the most noteworthy work, because it is not only the beginning of NEC speech recognition, also led to a long and productive research programs. In the 60 's, 3 research projects for more than 20 years of speech recognition research and development has had a significant impact. First research project in RCA lab started in the late 60 's, aims to voice the reality of events on the time scale is not uniform solutions. Reliably detect speech on the basis of the whole incident, developed a time conversion methods.

Speech recognition technology through joint efforts of scientists worldwide, more than half a century of research, now is close to a practical stage. In a lab environment, large-vocabulary continuous talk aloud broadband more than average recognition rate of speech signal can reach 90%. After having such high recognition rate, slowly from lab demonstration system of speech recognition technology towards practical items. IBM via voice and Dragon Dictation as a representative of the appearance of two dictation machine system so that the "speech recognition" coming into the limelight, has drawn wide attention.

Because of the trouble of proofreading and correcting recognition errors and a waste of time, thus reducing the advantage of voice recognition. Also, due to the use of accent or speech habits, the impact on the environment, speech recognition is greatly reduced, identify content not 100% correctly, so a lot of people believe that the speech recognition systems are unable to meet the practical requirements.

Currently, at&t and MIT, will voice recognition technology application in some limited vocabulary of specific task Shang, as phone automatically turned received, and phone query, and digital series recognition of task in the, dang speech of content is system by storage of content exists of, and using environment of acoustic characteristics and training data of acoustic characteristics difference not too big Shi, voice recognition of right recognition rate can close 100%. However, in practical use if these conditions are destroyed, then the system has a certain impact.

http://www.51due.net/writing/essay/sample30837.html

While speech recognition over the past 20 years there has been a lot of development, however, there is still a lot of shortcomings, needs to be further explored, can be classified into the following areas:

Speech-recognition technology can eliminate all kinds of acoustic environmental impact factors. In the noisy public environments, human consciousness will have to exclude non-acoustic environment factors are needed, the speech recognition system, it is difficult to do. In addition, in daily life, human language often has greater uncertainty, more casual, with clear speech habits. This, too, will give voice-recognition systems a lot of recognition problems. At present, in improving the reliability of voice systems in different environments, and to apply modern technology for voice recognition systems are more intelligent, mastering the language random part of the law, to achieve the best result.

System can identify the number of words is an important measure of what the system can do. A speech recognition system using acoustic model and voice model is too restrictive, when the user speaks of vocabulary beyond the known range of the system, the speech recognition system cannot accurately identify the appropriate content, for example, when suddenly shifted from Chinese to English, French, Russian, the computer will often mess odd output result. However, as system modeling method of continuous improvement, search algorithm to improve the efficiency of development and hardware resources, vocabulary speech recognition system may do so in the future without limits and languages mix, so that the user having to switch back and forth between languages when it is used, which can greatly reduce the restrictions on speech recognition system of vocabulary.

Voice recognition technology can for put fee brain, and effort, and time-consuming of machine operation into a is easy is has interesting of thing, like, dang people appeared hand busy, and hand cannot and the and avatar no operation of scene Shi, through voice recognition system of model structure, is can in as cab, and dangerous of industrial occasions, and far distance information gets, and appliances control, all aspects, voice recognition technology may led series new or more convenient function of equipment appeared, more convenient people of work and life. Its very broad scope of application and prospects. Not only can be used in daily life, revolution will bring more important means of production, is the basis for next-generation intelligent control.

Miniaturization is another important way to commercial application of speech recognition technology, extended drive depends on the degree itself advances in speech recognition technology, and closely related to the further development of the microelectronic chip technology. With advanced functionality and performance of speech recognition systems curing to more small chips or modules, can significantly lower the cost and volume of product by consumers of all ages, speech recognition systems and microelectronic chip technology development is the information technology revolution will lead us to a new level.

21st century is the era of information and networking, Internet and mobile communications, fixed-line telecommunications network connecting around the world. Natural spoken dialogue, e-commerce, information request, digital library, speech translation, distance education and other human activities that can be implemented on the network. The emergence of speech recognition systems, will let people communicate more freely, so people anywhere, any time, on anything to interact by voice, easy to enjoy more of the social modernization of information resources and services. This is a speech recognition technology research and application of important trends.

想要了解更多英国留学生新闻或者需要英国论文辅导，请关注51Due英国论文代写平台，51Due是一家专业的论文代写机构，专业辅导海外留学生的英文论文写作，主要业务有代写英国Essay、代写paper、代写assignment。在这里，51Due致力于为留学生朋友提供高效优质的留学教育辅导服务，为广大留学生提升写作水平，帮助他们达成学业目标。如果您有论文代写需求，可以咨询我们的客服QQ：800020041。

51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。-ZR