Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707704
Vimal Manohar, S. C. Bhargav, S. Umesh
In this paper, we propose a new acoustic modeling technique called the Phone-Cluster Adaptive Training. In this approach, the parameters of context-dependent states are obtained by the linear interpolation of several monophone cluster models, which are themselves obtained by adaptation using linear transformation of a canonical Gaussian Mixture Model (GMM). This approach is inspired from the Cluster Adaptive Training (CAT) for speaker adaptation and the Subspace Gaussian Mixture Model (SGMM). The parameters of the model are updated in an adaptive training framework. The interpolation vectors implicitly capture the phonetic context information. The proposed approach shows substantial improvement over the Continuous Density Hidden Markov Model (CDHMM) and a similar performance to that of the SGMM, while using significantly fewer parameters than both the CDHMM and the SGMM.
{"title":"Acoustic modeling using transform-based phone-cluster adaptive training","authors":"Vimal Manohar, S. C. Bhargav, S. Umesh","doi":"10.1109/ASRU.2013.6707704","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707704","url":null,"abstract":"In this paper, we propose a new acoustic modeling technique called the Phone-Cluster Adaptive Training. In this approach, the parameters of context-dependent states are obtained by the linear interpolation of several monophone cluster models, which are themselves obtained by adaptation using linear transformation of a canonical Gaussian Mixture Model (GMM). This approach is inspired from the Cluster Adaptive Training (CAT) for speaker adaptation and the Subspace Gaussian Mixture Model (SGMM). The parameters of the model are updated in an adaptive training framework. The interpolation vectors implicitly capture the phonetic context information. The proposed approach shows substantial improvement over the Continuous Density Hidden Markov Model (CDHMM) and a similar performance to that of the SGMM, while using significantly fewer parameters than both the CDHMM and the SGMM.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116081264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707759
Liang Lu, Arnab Ghoshal, S. Renals
Speech recognition systems normally use handcrafted pronunciation lexicons designed by linguistic experts. Building and maintaining such a lexicon is expensive and time consuming. This paper concerns automatically learning a pronunciation lexicon for speech recognition. We assume the availability of a small seed lexicon and then learn the pronunciations of new words directly from speech that is transcribed at word-level. We present two implementations for refining the putative pronunciations of new words based on acoustic evidence. The first one is an expectation maximization (EM) algorithm based on weighted finite state transducers (WFSTs) and the other is its Viterbi approximation. We carried out experiments on the Switchboard corpus of conversational telephone speech. The expert lexicon has a size of more than 30,000 words, from which we randomly selected 5,000 words to form the seed lexicon. By using the proposed lexicon learning method, we have significantly improved the accuracy compared with a lexicon learned using a grapheme-to-phoneme transformation, and have obtained a word error rate that approaches that achieved using a fully handcrafted lexicon.
{"title":"Acoustic data-driven pronunciation lexicon for large vocabulary speech recognition","authors":"Liang Lu, Arnab Ghoshal, S. Renals","doi":"10.1109/ASRU.2013.6707759","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707759","url":null,"abstract":"Speech recognition systems normally use handcrafted pronunciation lexicons designed by linguistic experts. Building and maintaining such a lexicon is expensive and time consuming. This paper concerns automatically learning a pronunciation lexicon for speech recognition. We assume the availability of a small seed lexicon and then learn the pronunciations of new words directly from speech that is transcribed at word-level. We present two implementations for refining the putative pronunciations of new words based on acoustic evidence. The first one is an expectation maximization (EM) algorithm based on weighted finite state transducers (WFSTs) and the other is its Viterbi approximation. We carried out experiments on the Switchboard corpus of conversational telephone speech. The expert lexicon has a size of more than 30,000 words, from which we randomly selected 5,000 words to form the seed lexicon. By using the proposed lexicon learning method, we have significantly improved the accuracy compared with a lexicon learned using a grapheme-to-phoneme transformation, and have obtained a word error rate that approaches that achieved using a fully handcrafted lexicon.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124769340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707711
Ali Orkan Bayer, G. Riccardi
Spoken language understanding (SLU) systems extract semantic information from speech signals, which is usually mapped onto concept sequences. The distribution of concepts in dialogues are usually sparse. Therefore, general models may fail to model the concept distribution for a dialogue and semantic models can benefit from adaptation. In this paper, we present an instance-based approach for on-line adaptation of semantic models. We show that we can improve the performance of an SLU system on an utterance, by retrieving relevant instances from the training data and using them for on-line adapting the semantic models. The instance-based adaptation scheme uses two different similarity metrics edit distance and n-gram match score on three different to-kenizations; word-concept pairs, words, and concepts. We have achieved a significant improvement (6% relative) in the understanding performance by conducting rescoring experiments on the n-best lists that the SLU outputs. We have also applied a two-level adaptation scheme, where adaptation is first applied to the automatic speech recognizer (ASR) and then to the SLU.
{"title":"On-line adaptation of semantic models for spoken language understanding","authors":"Ali Orkan Bayer, G. Riccardi","doi":"10.1109/ASRU.2013.6707711","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707711","url":null,"abstract":"Spoken language understanding (SLU) systems extract semantic information from speech signals, which is usually mapped onto concept sequences. The distribution of concepts in dialogues are usually sparse. Therefore, general models may fail to model the concept distribution for a dialogue and semantic models can benefit from adaptation. In this paper, we present an instance-based approach for on-line adaptation of semantic models. We show that we can improve the performance of an SLU system on an utterance, by retrieving relevant instances from the training data and using them for on-line adapting the semantic models. The instance-based adaptation scheme uses two different similarity metrics edit distance and n-gram match score on three different to-kenizations; word-concept pairs, words, and concepts. We have achieved a significant improvement (6% relative) in the understanding performance by conducting rescoring experiments on the n-best lists that the SLU outputs. We have also applied a two-level adaptation scheme, where adaptation is first applied to the automatic speech recognizer (ASR) and then to the SLU.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125444210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707696
Yangyang Shi, M. Larson, C. Jonker
Conventional n-gram language models are known for their limited ability to capture long-distance dependencies and their brittleness with respect to within-domain variations. In this paper, we propose a k-component recurrent neural network language model using curriculum learning (CL-KRNNLM) to address within-domain variations. Based on a Dutch-language corpus, we investigate three methods of curriculum learning that exploit dedicated component models for specific sub-domains. Under an oracle situation in which context information is known during testing, we experimentally test three hypotheses. The first is that domain-dedicated models perform better than general models on their specific domains. The second is that curriculum learning can be used to train recurrent neural network language models (RNNLMs) from general patterns to specific patterns. The third is that curriculum learning, used as an implicit weighting method to adjust the relative contributions of general and specific patterns, outperforms conventional linear interpolation. Under the condition that context information is unknown during testing, the CL-KRNNLM also achieves improvement over conventional RNNLM by 13% relative in terms of word prediction accuracy. Finally, the CL-KRNNLM is tested in an additional experiment involving N-best rescoring on a standard data set. Here, the context domains are created by clustering the training data using Latent Dirichlet Allocation and k-means clustering.
{"title":"K-component recurrent neural network language models using curriculum learning","authors":"Yangyang Shi, M. Larson, C. Jonker","doi":"10.1109/ASRU.2013.6707696","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707696","url":null,"abstract":"Conventional n-gram language models are known for their limited ability to capture long-distance dependencies and their brittleness with respect to within-domain variations. In this paper, we propose a k-component recurrent neural network language model using curriculum learning (CL-KRNNLM) to address within-domain variations. Based on a Dutch-language corpus, we investigate three methods of curriculum learning that exploit dedicated component models for specific sub-domains. Under an oracle situation in which context information is known during testing, we experimentally test three hypotheses. The first is that domain-dedicated models perform better than general models on their specific domains. The second is that curriculum learning can be used to train recurrent neural network language models (RNNLMs) from general patterns to specific patterns. The third is that curriculum learning, used as an implicit weighting method to adjust the relative contributions of general and specific patterns, outperforms conventional linear interpolation. Under the condition that context information is unknown during testing, the CL-KRNNLM also achieves improvement over conventional RNNLM by 13% relative in terms of word prediction accuracy. Finally, the CL-KRNNLM is tested in an additional experiment involving N-best rescoring on a standard data set. Here, the context domains are created by clustering the training data using Latent Dirichlet Allocation and k-means clustering.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121062119","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707768
Ankur Gandhe, Longlu Qin, Florian Metze, Alexander I. Rudnicky, Ian Lane, Matthias Eck
For low resource languages, collecting sufficient training data to build acoustic and language models is time consuming and often expensive. But large amounts of text data, such as online newspapers, web forums or online encyclopedias, usually exist for languages that have a large population of native speakers. This text data can be easily collected from the web and then used to both expand the recognizer's vocabulary and improve the language model. One challenge, however, is normalizing and filtering the web data for a specific task. In this paper, we investigate the use of online text resources to improve the performance of speech recognition specifically for the task of keyword spotting. For the five languages provided in the base period of the IARPA BABEL project, we automatically collected text data from the web using only Limited LP resources. We then compared two methods for filtering the web data, one based on perplexity ranking and the other based on out-of-vocabulary (OOV) word detection. By integrating the web text into our systems, we observed significant improvements in keyword spotting accuracy for four out of the five languages. The best approach obtained an improvement in actual term weighted value (ATWV) of 0.0424 compared to a baseline system trained only on LimitedLP resources. On average, ATWV was improved by 0.0243 across five languages.
{"title":"Using web text to improve keyword spotting in speech","authors":"Ankur Gandhe, Longlu Qin, Florian Metze, Alexander I. Rudnicky, Ian Lane, Matthias Eck","doi":"10.1109/ASRU.2013.6707768","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707768","url":null,"abstract":"For low resource languages, collecting sufficient training data to build acoustic and language models is time consuming and often expensive. But large amounts of text data, such as online newspapers, web forums or online encyclopedias, usually exist for languages that have a large population of native speakers. This text data can be easily collected from the web and then used to both expand the recognizer's vocabulary and improve the language model. One challenge, however, is normalizing and filtering the web data for a specific task. In this paper, we investigate the use of online text resources to improve the performance of speech recognition specifically for the task of keyword spotting. For the five languages provided in the base period of the IARPA BABEL project, we automatically collected text data from the web using only Limited LP resources. We then compared two methods for filtering the web data, one based on perplexity ranking and the other based on out-of-vocabulary (OOV) word detection. By integrating the web text into our systems, we observed significant improvements in keyword spotting accuracy for four out of the five languages. The best approach obtained an improvement in actual term weighted value (ATWV) of 0.0424 compared to a baseline system trained only on LimitedLP resources. On average, ATWV was improved by 0.0243 across five languages.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126622811","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707764
Y. Qian, Kai Yu, Jia Liu
Large vocabulary continuous speech recognition (LVCSR) is particularly difficult for low-resource languages, where only very limited manually transcribed data are available. However, it is often feasible to obtain large amount of untranscribed data of the low-resource target language or sufficient transcribed data of some non-target languages. Borrowing data from these additional sources to help LVCSR for low-resource language becomes an important research direction. This paper presents an integrated data borrowing framework in this scenario. Three data borrowing approaches were first investigated in detail, including feature, model and data corpus. They borrow data at different levels from additional sources, and all get substantial performance improvements. As these strategies work independently, the obtained gains are likely additive. The three strategies are then combined to form an integrated data borrowing framework. Experiments showed that with the integrated data borrowing framework, significant improvement of more than 10% absolute WER reduction over a conventional baseline was obtained. In particular, the gain under the extreme limited low-resource scenario is 16%.
{"title":"Combination of data borrowing strategies for low-resource LVCSR","authors":"Y. Qian, Kai Yu, Jia Liu","doi":"10.1109/ASRU.2013.6707764","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707764","url":null,"abstract":"Large vocabulary continuous speech recognition (LVCSR) is particularly difficult for low-resource languages, where only very limited manually transcribed data are available. However, it is often feasible to obtain large amount of untranscribed data of the low-resource target language or sufficient transcribed data of some non-target languages. Borrowing data from these additional sources to help LVCSR for low-resource language becomes an important research direction. This paper presents an integrated data borrowing framework in this scenario. Three data borrowing approaches were first investigated in detail, including feature, model and data corpus. They borrow data at different levels from additional sources, and all get substantial performance improvements. As these strategies work independently, the obtained gains are likely additive. The three strategies are then combined to form an integrated data borrowing framework. Experiments showed that with the integrated data borrowing framework, significant improvement of more than 10% absolute WER reduction over a conventional baseline was obtained. In particular, the gain under the extreme limited low-resource scenario is 16%.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126509334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707750
Pierre L. Dognin, V. Goel
Minimum phone error (MPE) training of deep neural networks (DNN) is an effective technique for reducing word error rate of automatic speech recognition tasks. This training is often carried out using a Hessian-free (HF) quasi-Newton approach, although other methods such as stochastic gradient descent have also been applied successfully. In this paper we present a novel stochastic approach to HF sequence training inspired by recently proposed stochastic average gradient (SAG) method. SAG reuses gradient information from past updates, and consequently simulates the presence of more training data than is really observed for each model update. We extend SAG by dynamically weighting the contribution of previous gradients, and by combining it to a stochastic HF optimization. We term the resulting procedure DSAG-HF. Experimental results for training DNNs on 1500h of audio data show that compared to baseline HF training, DSAG-HF leads to better held-out MPE loss after each model parameter update, and converges to an overall better loss value. Furthermore, since each update in DSAG-HF takes place over smaller amount of data, this procedure converges in about half the time as baseline HF sequence training.
{"title":"Combining stochastic average gradient and Hessian-free optimization for sequence training of deep neural networks","authors":"Pierre L. Dognin, V. Goel","doi":"10.1109/ASRU.2013.6707750","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707750","url":null,"abstract":"Minimum phone error (MPE) training of deep neural networks (DNN) is an effective technique for reducing word error rate of automatic speech recognition tasks. This training is often carried out using a Hessian-free (HF) quasi-Newton approach, although other methods such as stochastic gradient descent have also been applied successfully. In this paper we present a novel stochastic approach to HF sequence training inspired by recently proposed stochastic average gradient (SAG) method. SAG reuses gradient information from past updates, and consequently simulates the presence of more training data than is really observed for each model update. We extend SAG by dynamically weighting the contribution of previous gradients, and by combining it to a stochastic HF optimization. We term the resulting procedure DSAG-HF. Experimental results for training DNNs on 1500h of audio data show that compared to baseline HF training, DSAG-HF leads to better held-out MPE loss after each model parameter update, and converges to an overall better loss value. Furthermore, since each update in DSAG-HF takes place over smaller amount of data, this procedure converges in about half the time as baseline HF sequence training.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122545013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707772
Joris Driesen, S. Renals
Since subtitling television content is a costly process, there are large potential advantages to automating it, using automatic speech recognition (ASR). However, training the necessary acoustic models can be a challenge, since the available training data usually lacks verbatim orthographic transcriptions. If there are approximate transcriptions, this problem can be overcome using light supervision methods. In this paper, we perform speech recognition on broadcasts of Weatherview, BBC's daily weather report, as a first step towards automatic subtitling. For training, we use a large set of past broadcasts, using their manually created subtitles as approximate transcriptions. We discuss and and compare two different light supervision methods, applying them to this data. The best training set finally obtained with these methods is used to create a hybrid deep neural network-based recognition system, which yields high recognition accuracies on three separate Weatherview evaluation sets.
{"title":"Lightly supervised automatic subtitling of weather forecasts","authors":"Joris Driesen, S. Renals","doi":"10.1109/ASRU.2013.6707772","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707772","url":null,"abstract":"Since subtitling television content is a costly process, there are large potential advantages to automating it, using automatic speech recognition (ASR). However, training the necessary acoustic models can be a challenge, since the available training data usually lacks verbatim orthographic transcriptions. If there are approximate transcriptions, this problem can be overcome using light supervision methods. In this paper, we perform speech recognition on broadcasts of Weatherview, BBC's daily weather report, as a first step towards automatic subtitling. For training, we use a large set of past broadcasts, using their manually created subtitles as approximate transcriptions. We discuss and and compare two different light supervision methods, applying them to this data. The best training set finally obtained with these methods is used to create a hybrid deep neural network-based recognition system, which yields high recognition accuracies on three separate Weatherview evaluation sets.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122940070","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707748
Naoyuki Kanda, Ryu Takeda, Y. Obuchi
An acoustic model based on hidden Markov models with deep neural networks (DNN-HMM) has recently been proposed and achieved high recognition accuracy. In this paper, we investigated an elastic spectral distortion method to artificially augment training samples to help DNN-HMMs acquire enough robustness even when there are a limited number of training samples. We investigated three distortion methods - vocal tract length distortion, speech rate distortion, and frequency-axis random distortion - and evaluated those methods with Japanese lecture recordings. In a large vocabulary continuous speech recognition task with only 10 hours of training samples, a DNN-HMM trained with the elastic spectral distortion method achieved a 10.1% relative word error reduction compared with a normally trained DNN-HMM.
{"title":"Elastic spectral distortion for low resource speech recognition with deep neural networks","authors":"Naoyuki Kanda, Ryu Takeda, Y. Obuchi","doi":"10.1109/ASRU.2013.6707748","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707748","url":null,"abstract":"An acoustic model based on hidden Markov models with deep neural networks (DNN-HMM) has recently been proposed and achieved high recognition accuracy. In this paper, we investigated an elastic spectral distortion method to artificially augment training samples to help DNN-HMMs acquire enough robustness even when there are a limited number of training samples. We investigated three distortion methods - vocal tract length distortion, speech rate distortion, and frequency-axis random distortion - and evaluated those methods with Japanese lecture recordings. In a large vocabulary continuous speech recognition task with only 10 hours of training samples, a DNN-HMM trained with the elastic spectral distortion method achieved a 10.1% relative word error reduction compared with a normally trained DNN-HMM.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116978970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-12-01DOI: 10.1109/ASRU.2013.6707756
Kris Demuynck, Fabian Triefenbach
Deep neural networks (DNNs) have been shown to outperform Gaussian Mixture Models (GMM) on a variety of speech recognition benchmarks. In this paper we analyze the differences between the DNN and GMM modeling techniques and port the best ideas from the DNN-based modeling to a GMM-based system. By going both deep (multiple layers) and wide (multiple parallel sub-models) and by sharing model parameters, we are able to close the gap between the two modeling techniques on the TIMIT database. Since the `deep' GMMs retain the maximum-likelihood trained Gaussians as first layer, advanced techniques such as speaker adaptation and model-based noise robustness can be readily incorporated. Regardless of their similarities, the DNNs and the deep GMMs still show a sufficient amount of complementarity to allow effective system combination.
{"title":"Porting concepts from DNNs back to GMMs","authors":"Kris Demuynck, Fabian Triefenbach","doi":"10.1109/ASRU.2013.6707756","DOIUrl":"https://doi.org/10.1109/ASRU.2013.6707756","url":null,"abstract":"Deep neural networks (DNNs) have been shown to outperform Gaussian Mixture Models (GMM) on a variety of speech recognition benchmarks. In this paper we analyze the differences between the DNN and GMM modeling techniques and port the best ideas from the DNN-based modeling to a GMM-based system. By going both deep (multiple layers) and wide (multiple parallel sub-models) and by sharing model parameters, we are able to close the gap between the two modeling techniques on the TIMIT database. Since the `deep' GMMs retain the maximum-likelihood trained Gaussians as first layer, advanced techniques such as speaker adaptation and model-based noise robustness can be readily incorporated. Regardless of their similarities, the DNNs and the deep GMMs still show a sufficient amount of complementarity to allow effective system combination.","PeriodicalId":265258,"journal":{"name":"2013 IEEE Workshop on Automatic Speech Recognition and Understanding","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115567237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：