Speed improvement of the time-asynchronous acoustic fast match
نویسندگان
چکیده
This paper describes an algorithm for improvement of the speed of a time-asynchronous fast match, which is a part of a stack-search based recognition system. This fast match uses a phonetic tree to represent the entire vocabulary of the recognizer. Evaluation of the tree (in a depthrst manner), can be done much more e ciently using the fact that under certain conditions, the results of branch evaluations can be used to approximate the scores of other branches of the tree.
منابع مشابه
A Simulation Study of Multi-path Characteristics of Acoustic Propagation in the Strait of Hormuz
Multi-path interference due to boundary reflection in shallow water acoustic communication poses a major obstacle to reliable and high-speed underwater communication system. In this study, initially 3D variations of field data such as sound speed, temperature and salinity in horizontal transects of the Strait of Hormuz were analyzed using the ROPME data. Later, data on typical sound speed ...
متن کاملA New Verification-based Fast Match Approach to Large Vocabulary Constinuous Speech Recognition
Acoustic fast match is usually used to accelerate search in large vocabulary continuous speech recognition. This paper discusses a new acoustic fast match algorithm. This proposed fast match is based on incremental evaluation of the score and the use of normalized likelihood scores. This is in contrast to more traditional fast matches where a likelihood score is used. In addition, streaming SIM...
متن کاملA time-synchronous, tree-based search strategy in the acoustic fast match of an asynchronous speech recognition system
A common tactic in large vocabulary automatic speech recognition systems is to quickly provide short lists of likely candidate words from the vocabulary of several thousand possibilities. Subsequently, a detailed model for each of the words in a given short list is used to match that word to the acoustic signal. The process of determining the short list, called the fast match [1] reduces to a m...
متن کاملReducing search complexity in low perplexity tasks
In this paper we present a new method for improving the throughput of an asynchronous stack search based speech recognition system in the low perplexity applications. The algorithm reduces the acoustic fast match use in the cases where the word context information represented by the language model is sufcient to provide a reliable list of word candidates for the detailed match processing. The p...
متن کاملANFIS modeling and validation of a variable speed wind turbine based on actual data
In this research paper, ANFIS modeling and validation of Vestas 660 kW wind turbine based on actual data obtained from Eoun-Ebn-Ali wind farm in Tabriz, Iran, and FAST is performed. The turbine modeling is performed by deriving the non-linear dynamic equations of different subsystems. Then, the model parameters are identified to match the actual response. ANFIS is an artificial intelligent tech...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1999