scispace - formally typeset
Search or ask a question
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
More filters
Proceedings Article
01 Jan 1998
TL;DR: This paper presents an entropy-based algorithm for accurate and robust endpoint detection for speech recognition under noisy environments that uses the spectral entropy to identify the speech segments accurately.
Abstract: This paper presents an entropy-based algorithm for accurate and robust endpoint detection for speech recognition under noisy environments. Instead of using the conventional energy-based features, the spectral entropy is developed to identify the speech segments accurately. Experimental results show that this algorithm outperforms the energy-based algorithms in both detection accuracy and recognition performance under noisy environments, with an average error rate reduction of more than 16%.

221 citations

Journal ArticleDOI
TL;DR: It is argued that the written language notion of the word has had too much impact on models of spoken word recognition and a view of continuous word recognition is presented which takes into account the alternating pattern of weak and strong syllables in the speech stream.

221 citations

Proceedings ArticleDOI
19 Apr 2015
TL;DR: In this article, a multimodal learning approach was proposed for fusing speech and visual modalities for audio-visual automatic speech recognition (AV-ASR) using uni-modal deep networks.
Abstract: In this paper, we present methods in deep multimodal learning for fusing speech and visual modalities for Audio-Visual Automatic Speech Recognition (AV-ASR). First, we study an approach where uni-modal deep networks are trained separately and their final hidden layers fused to obtain a joint feature space in which another deep network is built. While the audio network alone achieves a phone error rate (PER) of 41% under clean condition on the IBM large vocabulary audio-visual studio dataset, this fusion model achieves a PER of 35.83% demonstrating the tremendous value of the visual channel in phone classification even in audio with high signal to noise ratio. Second, we present a new deep network architecture that uses a bilinear softmax layer to account for class specific correlations between modalities. We show that combining the posteriors from the bilinear networks with those from the fused model mentioned above results in a further significant phone error rate reduction, yielding a final PER of 34.03%.

220 citations

Journal ArticleDOI
TL;DR: A methodology for variable-star classification, drawing from modern machine-learning techniques, which is effective for identifying samples of specific science classes and presents the first astronomical use of hierarchical classification methods to incorporate a known class taxonomy in the classifier.
Abstract: With the coming data deluge from synoptic surveys, there is a growing need for frameworks that can quickly and automatically produce calibrated classification probabilities for newly-observed variables based on a small number of time-series measurements. In this paper, we introduce a methodology for variable-star classification, drawing from modern machine-learning techniques. We describe how to homogenize the information gleaned from light curves by selection and computation of real-numbered metrics ("feature"), detail methods to robustly estimate periodic light-curve features, introduce tree-ensemble methods for accurate variable star classification, and show how to rigorously evaluate the classification results using cross validation. On a 25-class data set of 1542 well-studied variable stars, we achieve a 22.8% overall classification error using the random forest classifier; this represents a 24% improvement over the best previous classifier on these data. This methodology is effective for identifying samples of specific science classes: for pulsational variables used in Milky Way tomography we obtain a discovery efficiency of 98.2% and for eclipsing systems we find an efficiency of 99.1%, both at 95% purity. We show that the random forest (RF) classifier is superior to other machine-learned methods in terms of accuracy, speed, and relative immunity to features with no useful class information; the RF classifier can also be used to estimate the importance of each feature in classification. Additionally, we present the first astronomical use of hierarchical classification methods to incorporate a known class taxonomy in the classifier, which further reduces the catastrophic error rate to 7.8%. Excluding low-amplitude sources, our overall error rate improves to 14%, with a catastrophic error rate of 3.5%.

217 citations

Journal ArticleDOI
TL;DR: A specific modeling methodology based on the study of errorcurves is introduced, which shows that CART decision-tree models providethe best estimation for default with an average 8.31% error rate, and the possibilities to use this type of accurate predictive model as ingredients of institutional and global risk models.
Abstract: Risk assessment of financial intermediaries is an area of renewed interest due to the financial crises of the 1980's and 90's. An accurate estimation of risk, and its use in corporate or global financial risk models, could be translated into a more efficient use of resources. One important ingredient to accomplish this goal is to find accurate predictors of individual risk in the credit portfolios of institutions. In this context we make a comparative analysis of different statistical and machine learning modeling methods of classification on a mortgage loan data set with the motivation to understand their limitations and potential. We introduced a specific modeling methodology based on the study of error curves. Using state-of-the-art modeling techniques we built more than 9,000 models as part of the study. The results show that CART decision-tree models provide the best estimation for default with an average 8.31% error rate for a training sample of 2,000 records. As a result of the error curve analysis for this model we conclude that if more data were available, approximately 22,000 records, a potential 7.32% error rate could be achieved. Neural Networks provided the second best results with an average error of 11.00%. The K-Nearest Neighbor algorithm had an average error rate of 14.95%. These results outperformed the standard Probit algorithm which attained an average error rate of 15.13%. Finally we discuss the possibilities to use this type of accurate predictive model as ingredients of institutional and global risk models.

217 citations


Network Information
Related Topics (5)
Deep learning
79.8K papers, 2.1M citations
88% related
Feature extraction
111.8K papers, 2.1M citations
86% related
Convolutional neural network
74.7K papers, 2M citations
85% related
Artificial neural network
207K papers, 4.5M citations
84% related
Cluster analysis
146.5K papers, 2.9M citations
83% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023271
2022562
2021640
2020643
2019633
2018528