Digital processing of speech signals

The application of dynamic time warping (DTW) to the analysis and monitoring of batch processes is presented. This dynamic-programming-based technique has been used in the area of speech recognition for the recognition of isolated and connected words. DTW has the ability to synchronize two trajectories by appropriately translating, expanding, and contracting localized segments within both trajectories to achieve a minimum distance between the trajectories. Batch processes often are characterized by unsynchronized trajectories, due to the presence of batch-to-batch disturbances and the existence of physical constraints. To compare these batch histories and apply statistical analysis one needs to reconcile the timing differences among these trajectories. This can be achieved using DTW with only a minimal amount of process knowledge. The combination of DTW and a monitoring method based on Multiway PCA/PLS is used for both off-line and on-line implementation. Data fiom an industrial polymerization reactor are used to illustrate the implementation and the performance of this method.

Synchronization of batch trajectories using dynamic time warping

The use of hidden Markov models (HMMs) in continuous speech recognition is reviewed. Markov models are presented as a generalization of their predecessor technology, dynamic programming. A unified view is offered in which both linguistic decoding and acoustic matching are integrated into a single, optimal network search framework. Advances in recognition architectures are discussed. The fundamentals of Viterbi beam search, the dominant search algorithm used today in speed recognition, are presented. Approaches to estimating the probabilities associated with an HMM model are examined. The HMM-supervised training paradigm is examined. Several examples of successful HMM-based speech recognition systems are reviewed. >

/pdf/continuous-speech-recognition-using-hidden-markov-models-2yiawm6wtz.pdf

Continuous speech recognition using hidden Markov models

The performance of two techniques is compared for automated recognition of bird song units from continuous recordings. The advantages and limitations of dynamic time warping (DTW) and hidden Markov models (HMMs) are evaluated on a large database of male songs of zebra finches (Taeniopygia guttata) and indigo buntings (Passerina cyanea), which have different types of vocalizations and have been recorded under different laboratory conditions. Depending on the quality of recordings and complexity of song, the DTW-based technique gives excellent to satisfactory performance. Under challenging conditions such as noisy recordings or presence of confusing short-duration calls, good performance of the DTW-based technique requires careful selection of templates that may demand expert knowledge. Because HMMs are trained, equivalent or even better performance of HMMs can be achieved based only on segmentation and labeling of constituent vocalizations, albeit with many more training examples than DTW templates. One weakness in HMM performance is the misclassification of short-duration vocalizations or song units with more variable structure (e.g., some calls, and syllables of plastic songs). To address these and other limitations, new approaches for analyzing bird vocalizations are discussed.

https://pondside.uchicago.edu/oba/Faculty/Margoliash/lab/pdfs/1997%20Kogan%20JASA.pdf

Automated recognition of bird song elements from continuous recordings using dynamic time warping and hidden Markov models: a comparative study.

Because of large variations involved in handwritten words, the recognition problem is very difficult. Hidden Markov models (HMM) have been widely and successfully used in speech processing and recognition. Recently HMM has also been used with some success in recognizing handwritten words with presegmented letters. In this paper, a complete scheme for totally unconstrained handwritten word recognition based on a single contextual hidden Markov model type stochastic network is presented. Our scheme includes a morphology and heuristics based segmentation algorithm, a training algorithm that can adapt itself with the changing dictionary, and a modified Viterbi algorithm which searches for the (l+1)th globally best path based on the previous l best paths. Detailed experiments are carried out and successful recognition results are reported. >

Off-line handwritten word recognition using a hidden Markov model type stochastic network

Principles of dynamic programming and its application to discrete-utterance and connected-speech recognition are introduced and discussed. The deterministic form, used for template matching for connected speech, is described in detail, and a number of algorithms are examined. The Viterbi algorithm, which is a form of dynamic programming for a stochastic system, is briefly discussed. >

The application of dynamic programming to connected speech recognition

It is often important that recognition decisions be made nearly instantaneously, i.e., early-decision, in applied connected-speech recognition systems. Therefore, many algorithms, which defer decisions until the end of an utterance, are unsuitable for these systems. This paper describes an early-decision recognizer, which has been operational at LEMS for about two years. For system-development purposes, the recognizer is programmed to reduce the data and make decisions sequentially using a multi-microprocessor implementation. While impractical for a development system, all algorithms are designed for a parallel, real-time, early-decision implementation. Connected-digit performance analysis, obtained from an automatic-evaluation system, is presented for ten male and ten female talkers. The reported evaluation has established a baseline word accuracy of 91.5%.

An early-decision, real-time, connected-speech recognizer

A facility for evaluating a talker-dependent, connected-speech recognition system is described. It is implemented as an independent system and interacts in parallel with a recognizer in real-time. The evaluator includes software for speech acquisition and storage, connected-speech training, data transfer to a recognizer, database queries, and statistical analysis. Important considerations in the design were the human factors of recording, talker and recording-condition variability, and the embedded training paradigm. Automatic statistical analysis is derived via a simple string-alignment algorithm using just the orthography. In order to demonstrate the use of this system, two experiments are described for connected-digit recognition. These results are presented as automatlcally-generated confusion matrices for insertion, substitution and deletion error and individual string alignments.

A real-time evaluation system for a real-time connected-speech recognizer

This paper analyzes a particular parallel architecture which incorporates eight Texas Instruments TMS32010 digital signal processors. The fully built and operational multiprocessor system, named PT8, is described, and several common DSP algorithms are presented which illustrate its characteristics. It is shown that data handling considerations are the major factors in the performance of these algorithms. Suggestions for upgrading this architecture to one which manages data more efficiently are given.

An investigation into the efficiency of a parallel TMS320 architecture: DFT and speech filterbank applications

Speech-source models indicate that event-synchronous feature analysis for a standard real-time speech recognizer (speaker trained using dynamic programming) should provide a strong basis for improved performance. Among the problems which need to be addressed in implementing such a system are those of developing robust event (pitch, burst, etc.) detectors, synchronous-analysis methodologies, more meaningful feature sets, and algorithms for recognition, e.g. dynamic programming. Descriptions of current approaches to these problems for a left-to-right, connected-word recognizer are given. The event-synchronous system's performance, relative to standard asynchronous recognition methodologies, is presented for an easy and a difficult vocabulary. >

D. Morgan

Papers

The application of dynamic programming to connected speech recognition

An early-decision, real-time, connected-speech recognizer

A real-time evaluation system for a real-time connected-speech recognizer

An investigation into the efficiency of a parallel TMS320 architecture: DFT and speech filterbank applications

An event-synchronous signal processing system for connected-speech recognition