<< 1 >>
Rating: 
Summary: A Confusing Book
Review: For studying the speech recognition subject this is not the right book to buy, It is hard to understand the theory using this book.
The c++ code works but there should be more remarks to make it easier to the readers to get along.
There is a free toolkit to download from the Internet named HTK
that contains full C code and a free book to the same theory so actually it is a waste of money to buy this book !!!
Rating: 
Summary: Good Starter but Confusing.
Review: I appreciated the balance between theory and implementation in the book. Also the content covers the most important topics. It is unfortunate that the book contains numerous typo's and confusing choices of symbols. The errors are often right in the most critical places too. The explanation of the theory of HMM's for example. 0 and 1 are chosen to represent both white and black balls and 2 different urns all in the same diagram. When trying to sort out which is which the reader will be further confused by blatant errors where a 0 should be a 1. I am afraid many new readers will find frustration on the theory sections.The choice of C++ and inclusion of a CD-ROM with full source is a nice touch however. Just be aware that the code is not geared for real-time recognition.
Rating:
Summary: Good Starter but Confusing.
Review: I appreciated the balance between theory and implementation in the book. Also the content covers the most important topics. It is unfortunate that the book contains numerous typo's and confusing choices of symbols. The errors are often right in the most critical places too. The explanation of the theory of HMM's for example. 0 and 1 are chosen to represent both white and black balls and 2 different urns all in the same diagram. When trying to sort out which is which the reader will be further confused by blatant errors where a 0 should be a 1. I am afraid many new readers will find frustration on the theory sections. The choice of C++ and inclusion of a CD-ROM with full source is a nice touch however. Just be aware that the code is not geared for real-time recognition.
Rating: 
Summary: speech recognition c
Review: I want to know about HMM. And c-source code
Rating:
Summary: please send book speech recognition
Review: please send book speech recognitio
Rating:
Summary: very good book if you read its code
Review: This book is composed of two parts, theory and implementation. if you only read its theoretical part, it is ok with many details missing. it is not clearly written. however, if you study its C++ code, you would get all you want on recognition system. I spent 3-8 hours everyday for 4 months going through its code line by line. The C++ code (30,000 lines in total) is very well written but without comments. Many times, I need to figure out things not written in the book. I once spent 1 week on 200 lines of code. However, After 4 months, I truely understand the system. You will find this book useful only if you really spend time covering its C++ code line by line. If you want theory only, goto read other books. I rate 2 star for its theory and 5 star for its implementation. [website]
Rating: 
Summary: Useful application-oriented book.
Review: This book is very useful for persons willing to implement a speech recognition system based on hidden Markov models (HMM). The authors provide the source code of a complete system. Each chapter is divided in two parts: theory and implementation. Some implementation issues are of interest only for those who develop code in C. I think this implementation-oriented book is a good complement for a theory-oriented book, as "Fundamentals of Speech Recog." by Rabiner and Juang. I really couldn't understand an appendix about "Econometric", that mainly discusses HREH (?) and just mentions HMM in the very end. The references related to this appendix were mixed with the ones related to speech, with an annoying result.
<< 1 >>
| 
