Topic
Word error rate
About: Word error rate is a research topic. Over the lifetime, 11939 publications have been published within this topic receiving 298031 citations.
Papers published on a yearly basis
Papers
More filters
TL;DR: The CSWC-SVM algorithm constructed in this study performs well in response to the attack of the industrial control system in the virtual reality simulation environment, which provides practical significance for the application of virtual reality in industrial monitoring.
Abstract: In order to protect industrial safety, improve the operation stability of the industrial control system, conduct the response measures for network environment attacked by the external world, and realize simulation in virtual reality environment, in this study, class and sample weighted C-support vector machine (CSWC-SVM) algorithm is first proposed using SVM. Then, the intrusion detection model of industrial control network is built based on the CSWC-SVM algorithm. Finally, KDD CUP 1999 data are introduced to carry out simulation experiments on the algorithm model constructed in this study in the virtual reality simulation environment. The results show when the penalty factor of the polynomial kernel function, radial basis kernel function, and sigmoid kernel function is 104, the average number of support vectors is 45, 46, and 37, respectively; the average training time are about 0.43, 0.45, and 0.47 s, and the average test time is about 9.7, 9.9, and 10.2 s, respectively; the average recognition accuracy is about 85.7%, 86.2%, and 86.7%, and the false positive rate is 3.8%, 2.8%, and 2.3%, respectively; the accuracy of the CSWC-SVM algorithm in different sample sizes (1000–6000) can be kept above 90%. The operation error rate of the CSWC-SVM algorithm is lower than that of C-SVM, C-SVM, and RS-SVM algorithms under different validation data sets. After dimension reduction, the classification accuracy of the CSWC-SVM algorithm is higher than that of C-SVM and WC-SVM algorithms. The weight value increases from 0 to 200, and the number of model errors on 1000, 2000, and 3000 pieces of data decreases significantly. When the weight value is 200, the number of errors drops to 0, and the classification accuracy reaches 100%. In a word, the CSWC-SVM algorithm constructed in this study performs well in response to the attack of the industrial control system in the virtual reality simulation environment, which provides practical significance for the application of virtual reality in industrial monitoring.
80 citations
14 Dec 2019
TL;DR: In this article, a system that generates speaker-annotated transcripts of meetings by using a microphone array and a 360-degree camera is described, which can handle overlapped speech.
Abstract: This paper describes a system that generates speaker-annotated transcripts of meetings by using a microphone array and a 360-degree camera. The hallmark of the system is its ability to handle overlapped speech, which has been an unsolved problem in realistic settings for over a decade. We show that this problem can be addressed by using a continuous speech separation approach. In addition, we describe an online audio-visual speaker diarization method that leverages face tracking and identification, sound source localization, speaker identification, and, if available, prior speaker information for robustness to various real world challenges. All components are integrated in a meeting transcription framework called SRD, which stands for “separate, recognize, and diarize”. Experimental results using recordings of natural meetings involving up to 11 attendees are reported. The continuous speech separation improves a word error rate (WER) by 16.1% compared with a highly tuned beamformer. When a complete list of meeting attendees is available, the discrepancy between WER and speaker-attributed WER is only 1.0%, indicating accurate word-to-speaker association. This increases marginally to 1.6% when 50% of the attendees are unknown to the system.
80 citations
17 Jul 2006
TL;DR: Substantial improvement in word-alignment accuracy is demonstrated, partly though improved training methods, but predominantly through selection of more and better features.
Abstract: For many years, statistical machine translation relied on generative models to provide bilingual word alignments. In 2005, several independent efforts showed that discriminative models could be used to enhance or replace the standard generative approach. Building on this work, we demonstrate substantial improvement in word-alignment accuracy, partly though improved training methods, but predominantly through selection of more and better features. Our best model produces the lowest alignment error rate yet reported on Canadian Hansards bilingual data.
79 citations
TL;DR: The results suggest that the minimum adequate number of training samples is at least ten times the number of classifier parameters.
79 citations
14 Apr 1991
TL;DR: A corrective MMIE training algorithm is introduced, which, when applied to the TI/NIST connected digit database, has made it possible to reduce the string error rate by close to 50%.
Abstract: Recently, Gopalakrishnan et al (1989) introduced a reestimation formula for discrete HMMs (hidden Markov models) which applies to rational objective functions like the MMIE (maximum mutual information estimation) criterion The authors analyze the formula and show how its convergence rate can be substantially improved They introduce a corrective MMIE training algorithm, which, when applied to the TI/NIST connected digit database, has made it possible to reduce the string error rate by close to 50% Gopalakrishnan's result is extended to the continuous case by proposing a new formula for estimating the mean and variance parameters of diagonal Gaussian densities >
79 citations