Can anyone help me understand HMM please, I have several questions... 1.How do you derive/get the transition probabilities of each state in HMM? 2.Why do you have to apply gauss function to each recognition object? 3.PLease give me the big picture of HMM cause I can't seem to really understand it fully(technically) although I have read about it so many times. WHat I understand only is you have to train a model for each recognition object(i.e. phoneme, word). Each model consists of states in which each transition either causes the occurence of a symbol or is caused by a symbol occurence which I'm not very sure about. Well that's all I know, I don't know how to get the probabilities and the other technical stuff. Please help me I'm so confused. ty

(You need to be a member of speech-recognition -- send a blank email to speech-recognition-subscribe@yahoogroups.com )