Language consists of sounds, right?. Every individual generates a unique set of sounds, correct?. However, one can distinguish standard speakers (SS), generating average kinds of sounds, who are subdivided according to their age, gender, region, social status, education, occupation etc. One should determine sound frequencies, probabilities of sounds following each other, intonation contours, vocabularies, physical properties of individual sounds for every standard speaker. That is not easy, but based on these data one can simulate natural speech flow! System, however, should also include statistic information about the population structure and with its help generate speech flows with those features, which characterize population of some region or the whole country. Frankly speaking, statistical model contains statistics data about the population structure, speech based on standard speakers, speech signal processing facilities (algorithms for synthesis), means of sounds parameters speed determination, algorithms for sounds generation and standard speakers distributions. Confused? Sounds tuff? You can read more and inquire an audio file with a speech model right from here (http://www.sevana.fi/audio_speech_codecs_quality_analysis.php). Next time we discuss what is Signals synchronizer and after that we’ll approach the core of this technology and its major know-how – the analytical module! See you tomorrow!
What is that statistical speech model you are talking about?
