Speech recognition HMM training on reconfigurable parallel processor
نویسندگان
چکیده
Armstmng III is a 20 node muiti-coqtwter that is current& operational In ah&ion to a RISC pmcessor, each node contains recoflfgurabfe *sources impiemented I&I FPGAs. The in-circuit rtpmgramabihy of static RAM-based FPGAs allows the computational crrpabilities of a node to be @amica& matcbed to the conq%etati~nal requirements of an @&cation. Most reconfgurable conguters in exiitence today re& so@ on a large number of FPGAs to petform coqtmtations. In contrast, thispaper ciemonstrates the uti@ of a smaI/ number of FPGAs coupled to a RISC pmcessor tvith a simple interconnect. This paper &scribes a substantive exa+/e qt@cation that peforms HMM Training for speech recognition n&b the reconjgurable phgonn.
منابع مشابه
Improving Phoneme Sequence Recognition using Phoneme Duration Information in DNN-HSMM
Improving phoneme recognition has attracted the attention of many researchers due to its applications in various fields of speech processing. Recent research achievements show that using deep neural network (DNN) in speech recognition systems significantly improves the performance of these systems. There are two phases in DNN-based phoneme recognition systems including training and testing. Mos...
متن کاملA new hybrid structure of speech recognizer based on HMM and neural network
In this paper, we introduced a new framework of speech recognizer based on HMM and neural net. Unlike the traditional hybrid system, the neural net was used as a post processor, which classify the speech data segmented by HMM recognizer. The purpose of this method is to improve the top-choice accuracy of HMM based speech recognition system in our lab. Major issues such as how to use the segment...
متن کاملDual-ALU Processor for Speech Signal Processing
This paper proposed a low-cost dual-ALU processor for speech recognition. It does not give an emphasis on sophistication but on low-cost solution. The dual-ALU architecture provides parallel calculation capability. For the consideration of chip size, the area of the second ALU is only half of the first ALU. We use hardware-software codesign method to implement the speech recognition. The featur...
متن کاملResearch in Continuous Speech Recognition
The primary goal of this work is to develop improved methods and models for acoustic recognition of continuous speech. Most of the work has focused on deriving statistical models for speech recognition that can capture the acoustic-phonetic phenomena that occur in speech with the constraint that the models can be adequately est imated from a reasonable amount of training speech. Most of our wor...
متن کاملThe use of meta-HMM in multistream HMM training for automatic speech recognition
Among the di erent attempts to improve recognition scores and robustness to noise, the recognition of parallel streams of data each one representing partial information on the test signal and the fusion of the decisions have received a great deal of interest. The problem of training such models taking recombination constraints at the level of speech-subunits has not yet been rigorously addresse...
متن کامل