This article has been key issues in the automatic transcription of speech automatic review ways to make a simple introduction. In the letter, the language of text to speech (text to speech, TTS) and automatic speech recognition systems (automatic speech recognition, speech recognition) system, the practical application, due to the emergence of new words and many proper nouns collection, storage difficulties pronunciation dictionary does not exist is often encountered, need these words automatic transcription. On this basis, gives an automated voice system implementation, impact on performance, some technical details of the analysis system and method are given hand-written on the comparison of experimental results show that under the rules, significantly better than the latter.

voice conversion is based on the combination of proposed independent target speaker pitch between the source of the transfer Because, although the conversion is in the field by modifying the residual signal IoSE Time interval to achieve, it is not on cloud / unvoiced decisions need And make the same silent treatment, which, in the expression part of the, IoSE Equivalent pitch information, and in the silent period.

Based on fuzzy mathematics gained considerable development in the theory and application have achieved fruitful results impressive. Applications of fuzzy mathematics has been related to automatic control, image and text recognition, artificial intelligence, geology, seismology, medical diagnostics, weather analysis, aviation, aerospace, ship train car to drive, traffic management, decision evaluation, business management and socio-economic many aspects.

This paper analyzes the principle of hierarchical reconstruction on the impact of the robot image recognitions given to the affine point from the image reconstruction of the corresponding instance of calculated measures, which to some extent is a measure of the corresponding affine point reconstruction needs of the most basic structural information, such as the measure of similar objects based on reconstruction, metric reconstruction of specular reflection based, reconstruction based on the measurement of moving objects, and based on hidden consumption point (line) of the metric reconstruction to improve image clarity.

Voice conversion is based on the combination of proposed independent target speaker pitch between the sources of the transfer information systems. Because, although the conversion is in the field by modifying the residual signal oise time interval to achieve, it is not on cloud/unvoiced decisions need and make the same silent treatment, which, in the expression part of the, IoSE equivalent pitch information, and in the silent period.

The Text Analysis Conference (TAC) is organized by the U.S. National Institute of Standards and Technology (NIST).

Our method to Knowledge Base Population at TAC2011 is described in this paper. Rule-based method and machine learning method are combined in the Slot Filling task. And in the Entity Linking task, query expansion method, rule-based method and entity disambiguation method are mainly used.

Data compression takes a very important role in the computer data storage applications and embedded systems. In this article, Huffman algorithm is put forward as popular lossless data compression method and an in-depth study. Some amelioration in Huffman algorithm is used on the theory, and applied on embedded systems (sitronixST2221C) features as actual demand, considering the data compression of some important indicators, such as: compression ratio, operation speed, the algorithm, the complexity to find a suitable compression algorithm, the optimized and applied to new products.

Today,robots are becoming more and more intelligent and humane.The interaction is the most effective,most commonly used and most convenient form of information exchange between robots and humans.The design of interactive robots is an inevitable trend in scientific development.

Through the binary processing on images,we can use median filter and average value filter manners to denoise the bar codes,solve the question which bar code meets in the recognition process.

This article attempts to carry on the classification to the result of University Personality Invention using artificial neural network.It is also an example of application and design of BP algorithm in UPI.