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...

The problem addressed by this paper is to enhance the continuous speech recognizers robustness to noise. For this purpose, the acoustic signal is filtered into several spectral bands, and independent recognition is achieved in each band. Then, the system recombines the results given by each recognizer and delivers a unique solution. The main advantage of this method is to consider the signal on...

Continuous speech production is a highly complex process involving many parts of the human brain. To date, no fundamental representation that allows for decoding of continuous speech from neural signals has been presented. Here we show that techniques from automatic speech recognition can be applied to decode a textual representation of spoken words from neural signals. We model phones as the f...

Large vocabulary speaker-independent speech recognition systems being capable of recognizing continuous speech based on hidden Markov models are today’s standard. This review introduces the fundamentals of speech and the underlying speech recognition problems. The three classical approaches, i.e., the acoustic-phonetic, the statistical (pattern) recognition and the artificial intelligence appro...

Continuous speech is far more natural and ecient than isolated speech for communication. However, for current state-of-the-art automatic speech recognition systems, isolated speech recognition (ISR) is far more accurate than continuous speech recognition (CSR). It is common practice in the speech research community to build CSR systems using only CS data. However, slowing of the speaking rate ...

Performance of any continuous speech recognition system is highly dependent on performance of the acoustic models. Generally, development of the robust spoken language technology relies on the availability of large amounts of data. Common way to cope with little data for training each state of Markov models is treebased state tying. This tying method applies contextual questions to tie states. ...

In this paper we describe the development of a doctor-patient dialogue corpus to support a speech-to-speech machine translation effort for English-Persian medical dialogues. The corpus was developed by recording and transcribing English-to-English dialogues between medical students and standardized patients (actors who have been trained to portray illness or injury victims), and then translated...

Optical Character Recognition (OCR) is an area of research that has attracted the interest of researchers for the past forty years. Although the subject has been the center topic for many researchers for years, it remains one of the most challenging and exciting areas in pattern recognition. Because of the cursive nature of Persian language, recognition of its characters is more difficult than ...

It is well known that the application of hidden Markov models (HMMs) has led to a dramatic increase of the performance of automatic speech recognition in the 1980s and from that time onwards. In particular, large vocabulary continuous speech recognition (LVCSR) could be realized by using a recognition unit such as phones. A variety of speech characteristics can be modelled by using HMMs effecti...

We propose a low-memory-bandwidth, high-efficiency VLSI architecture for 60-k word real-time continuous speech recognition. Our architecture includes a cache architecture using the locality of speech recognition, beam pruning using a dynamic threshold, two-stage language model searching, a parallel Gaussian Mixture Model (GMM) architecture based on the mixture level and frame level, a parallel ...