Pub Date : 2019-05-12DOI: 10.1109/ICASSP.2019.8683820
F. Bovino
Bell measurements, jointly projecting two qubits onto the so-called Bell basis, constitute a crucial step in many quantum computation and communication protocols, including dense coding, quantum repeaters, and teleportation-based quantum computation. A problem is the impossibility of deterministic unambiguous Bell measurements using passive linear optics, even when arbitrarily many auxiliary photons, photon-number-resolving detectors, and dynamical (conditionally changing) networks are available. Current proposals for going over the 50% upper bound without using experimentally challenging nonlinearities rely on using entangled photon ancilla states and a sufficiently large interferometer to combine the signal and ancilla modes. We demonstrate that the novel Multiple Rail architecture, based on the propagation of a single photon in a complex multipath optical circuit (or multiwaveguide optical circuit), provides the possibility to perform deterministic Bell measurements so to unambiguously discrimate all four Bell States.
{"title":"Intrasystem Entanglement Generator and Unambiguos Bell States Discriminator on Chip","authors":"F. Bovino","doi":"10.1109/ICASSP.2019.8683820","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8683820","url":null,"abstract":"Bell measurements, jointly projecting two qubits onto the so-called Bell basis, constitute a crucial step in many quantum computation and communication protocols, including dense coding, quantum repeaters, and teleportation-based quantum computation. A problem is the impossibility of deterministic unambiguous Bell measurements using passive linear optics, even when arbitrarily many auxiliary photons, photon-number-resolving detectors, and dynamical (conditionally changing) networks are available. Current proposals for going over the 50% upper bound without using experimentally challenging nonlinearities rely on using entangled photon ancilla states and a sufficiently large interferometer to combine the signal and ancilla modes. We demonstrate that the novel Multiple Rail architecture, based on the propagation of a single photon in a complex multipath optical circuit (or multiwaveguide optical circuit), provides the possibility to perform deterministic Bell measurements so to unambiguously discrimate all four Bell States.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"6 1","pages":"7993-7997"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74191445","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8683841
Matthieu Simeoni, P. Hurley
The starting point for deconvolution methods in radio astronomy is an estimate of the sky intensity called a dirty image. These methods rely on the telescope point-spread function so as to remove artefacts which pollute it. In this work, we show that the intensity field is only a partial summary statistic of the matched filtered interferometric data, which we prove is spatially correlated on the celestial sphere. This allows us to define a sky covariance function. This previously unexplored quantity brings us additional information that can be leveraged in the process of removing dirty image artefacts. We demonstrate this using a novel unsupervised learning method. The problem is formulated on a graph: each pixel interpreted as a node, linked by edges weighted according to their spatial correlation. We then use spectral clustering to separate the artefacts in groups, and identify physical sources within them.
{"title":"Graph Spectral Clustering of Convolution Artefacts in Radio Interferometric Images","authors":"Matthieu Simeoni, P. Hurley","doi":"10.1109/ICASSP.2019.8683841","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8683841","url":null,"abstract":"The starting point for deconvolution methods in radio astronomy is an estimate of the sky intensity called a dirty image. These methods rely on the telescope point-spread function so as to remove artefacts which pollute it. In this work, we show that the intensity field is only a partial summary statistic of the matched filtered interferometric data, which we prove is spatially correlated on the celestial sphere. This allows us to define a sky covariance function. This previously unexplored quantity brings us additional information that can be leveraged in the process of removing dirty image artefacts. We demonstrate this using a novel unsupervised learning method. The problem is formulated on a graph: each pixel interpreted as a node, linked by edges weighted according to their spatial correlation. We then use spectral clustering to separate the artefacts in groups, and identify physical sources within them.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"27 1","pages":"4260-4264"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74225932","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682418
Seokhwan Kim
Punctuation restoration is a post-processing task of automatic speech recognition to generate the punctuation marks on un-punctuated transcripts. This paper proposes a deep recurrent neural network architecture with layer-wise multi-head attentions towards better modelling of the contexts from a variety of perspectives in putting punctuations by human writers. The experimental results show that our proposed model significantly outperforms previous state-of-the-art methods in punctuation restoration performances on IWSLT dataset.
{"title":"Deep Recurrent Neural Networks with Layer-wise Multi-head Attentions for Punctuation Restoration","authors":"Seokhwan Kim","doi":"10.1109/ICASSP.2019.8682418","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682418","url":null,"abstract":"Punctuation restoration is a post-processing task of automatic speech recognition to generate the punctuation marks on un-punctuated transcripts. This paper proposes a deep recurrent neural network architecture with layer-wise multi-head attentions towards better modelling of the contexts from a variety of perspectives in putting punctuations by human writers. The experimental results show that our proposed model significantly outperforms previous state-of-the-art methods in punctuation restoration performances on IWSLT dataset.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"28 1","pages":"7280-7284"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74555551","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8683594
Lorenzo Dall'Amico, Romain Couillet
This article improves over the recently proposed Bethe Hessian matrix for community detection on sparse graphs, assuming here a more realistic setting where node degrees are inhomogeneous. We notably show that the parametrization proposed in the seminal work on the Bethe Hessian clustering can be ameliorated with positive consequences on correct classification rates. Extensive simulations support our claims.
{"title":"Community Detection in Sparse Realistic Graphs: Improving the Bethe Hessian","authors":"Lorenzo Dall'Amico, Romain Couillet","doi":"10.1109/ICASSP.2019.8683594","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8683594","url":null,"abstract":"This article improves over the recently proposed Bethe Hessian matrix for community detection on sparse graphs, assuming here a more realistic setting where node degrees are inhomogeneous. We notably show that the parametrization proposed in the seminal work on the Bethe Hessian clustering can be ameliorated with positive consequences on correct classification rates. Extensive simulations support our claims.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"61 1","pages":"2942-2946"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74633205","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8683066
Stefan Braun, Shih-Chii Liu
Recent work on neural networks with probabilistic parameters has shown that parameter uncertainty improves network regularization. Parameter-specific signal-to-noise ratio (SNR) levels derived from parameter distributions were further found to have high correlations with task importance. However, most of these studies focus on tasks other than automatic speech recognition (ASR). This work investigates end-to-end models with probabilistic parameters for ASR. We demonstrate that probabilistic networks outperform conventional deterministic networks in pruning and domain adaptation experiments carried out on the Wall Street Journal and CHiME-4 datasets. We use parameter-specific SNR information to select parameters for pruning and to condition the parameter updates during adaptation. Experimental results further show that networks with lower SNR parameters (1) tolerate increased sparsity levels during parameter pruning and (2) reduce catastrophic forgetting during domain adaptation.
{"title":"Parameter Uncertainty for End-to-end Speech Recognition","authors":"Stefan Braun, Shih-Chii Liu","doi":"10.1109/ICASSP.2019.8683066","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8683066","url":null,"abstract":"Recent work on neural networks with probabilistic parameters has shown that parameter uncertainty improves network regularization. Parameter-specific signal-to-noise ratio (SNR) levels derived from parameter distributions were further found to have high correlations with task importance. However, most of these studies focus on tasks other than automatic speech recognition (ASR). This work investigates end-to-end models with probabilistic parameters for ASR. We demonstrate that probabilistic networks outperform conventional deterministic networks in pruning and domain adaptation experiments carried out on the Wall Street Journal and CHiME-4 datasets. We use parameter-specific SNR information to select parameters for pruning and to condition the parameter updates during adaptation. Experimental results further show that networks with lower SNR parameters (1) tolerate increased sparsity levels during parameter pruning and (2) reduce catastrophic forgetting during domain adaptation.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"101 1","pages":"5636-5640"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79373113","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682318
Jiro Abe, M. Yamagishi, I. Yamada
In this paper, we propose a new linearly involved convexity-preserving model for signal recovery by extending the idea in the generalized minimax concave (GMC) penalty [Se-lesnick’ 17]. The proposed model can use nonconvex penalties but maintain the overall convexity and is applicable to much more general scenarios of signal recovery than the original GMC model. We also propose a new iterative algorithm which has theoretical guarantee of convergence to a global minimizer of the proposed model. A numerical experiment for noise suppression shows excellent edge-preserving performance of the proposed smoother in comparison with the standard convex TV smoother.
{"title":"Convexity-edge-preserving Signal Recovery with Linearly Involved Generalized Minimax Concave Penalty Function","authors":"Jiro Abe, M. Yamagishi, I. Yamada","doi":"10.1109/ICASSP.2019.8682318","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682318","url":null,"abstract":"In this paper, we propose a new linearly involved convexity-preserving model for signal recovery by extending the idea in the generalized minimax concave (GMC) penalty [Se-lesnick’ 17]. The proposed model can use nonconvex penalties but maintain the overall convexity and is applicable to much more general scenarios of signal recovery than the original GMC model. We also propose a new iterative algorithm which has theoretical guarantee of convergence to a global minimizer of the proposed model. A numerical experiment for noise suppression shows excellent edge-preserving performance of the proposed smoother in comparison with the standard convex TV smoother.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"5 1","pages":"4918-4922"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85243683","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682498
Camelia Elisei-Iliescu, C. Paleologu, J. Benesty, S. Ciochină
The recursive least-squares (RLS) adaptive filter is an appealing choice in system identification problems, mainly due to its fast convergence rate. However, this algorithm is computationally very complex, which may make it useless for the identification of high length impulse responses, like in echo cancellation. In this paper, we focus on a new approach to improve the efficiency of the RLS algorithm. The basic idea is to exploit the impulse response decomposition based on the nearest Kronecker product and low-rank approximation. Thus, a high-dimension system identification problem is reformulated in terms of low-dimension problems, which are tensorized together. Simulations performed in the context of echo cancellation indicate the good performance of the RLS algorithm based on this approach.
{"title":"A Recursive Least-squares Algorithm Based on the Nearest Kronecker Product Decomposition","authors":"Camelia Elisei-Iliescu, C. Paleologu, J. Benesty, S. Ciochină","doi":"10.1109/ICASSP.2019.8682498","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682498","url":null,"abstract":"The recursive least-squares (RLS) adaptive filter is an appealing choice in system identification problems, mainly due to its fast convergence rate. However, this algorithm is computationally very complex, which may make it useless for the identification of high length impulse responses, like in echo cancellation. In this paper, we focus on a new approach to improve the efficiency of the RLS algorithm. The basic idea is to exploit the impulse response decomposition based on the nearest Kronecker product and low-rank approximation. Thus, a high-dimension system identification problem is reformulated in terms of low-dimension problems, which are tensorized together. Simulations performed in the context of echo cancellation indicate the good performance of the RLS algorithm based on this approach.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"1 1","pages":"4843-4847"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85421440","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682469
Tianwen Wei, S. Chrétien
We propose Independent Component Autoencoder (ICAE), a deep neural network-based framework for nonlinear Independent Component Analysis (ICA). The proposed method consists of a penalized autoencoder and a training objective that is to minimize a combination of the reconstruction loss and an ICA contrast. Unlike many previous ICA methods that are usually tailored to separate specific mixture, our method can recover sources from various mixtures, without prior knowledge on the nature of that mixture.
{"title":"A Penalized Autoencoder Approach for Nonlinear Independent Component Analysis","authors":"Tianwen Wei, S. Chrétien","doi":"10.1109/ICASSP.2019.8682469","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682469","url":null,"abstract":"We propose Independent Component Autoencoder (ICAE), a deep neural network-based framework for nonlinear Independent Component Analysis (ICA). The proposed method consists of a penalized autoencoder and a training objective that is to minimize a combination of the reconstruction loss and an ICA contrast. Unlike many previous ICA methods that are usually tailored to separate specific mixture, our method can recover sources from various mixtures, without prior knowledge on the nature of that mixture.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"51 1","pages":"2797-2801"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79813638","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682368
Yanfeng Lu, M. Dong, Ying Chen
Text-to-Speech (TTS) systems have been evolving rapidly in recent years. With the great modelling power of deep neural networks, researchers have achieved end-to-end conversion from raw text to speech. It has been shown by various research projects that end-to-end TTS systems are able to generate speech that sounds akin to human voice for English and other languages. However, for languages like Chinese, there are two problems to deal with. Firstly, due to the large character set, a small input set comparable to the English character set is needed for the end-to-end solution. Secondly, there are serious prosodic phrasing mistakes when the end-to-end method is applied to Chinese. In this paper, we will propose a solution for an end-to-end Chinese TTS system on the basis of Tacotron 2 and Wavenet vocoder. We will then add extra contextual information to improve the performance of prosodic phrasing. Our experiments have demonstrated the effectiveness of this proposal.
{"title":"Implementing Prosodic Phrasing in Chinese End-to-end Speech Synthesis","authors":"Yanfeng Lu, M. Dong, Ying Chen","doi":"10.1109/ICASSP.2019.8682368","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682368","url":null,"abstract":"Text-to-Speech (TTS) systems have been evolving rapidly in recent years. With the great modelling power of deep neural networks, researchers have achieved end-to-end conversion from raw text to speech. It has been shown by various research projects that end-to-end TTS systems are able to generate speech that sounds akin to human voice for English and other languages. However, for languages like Chinese, there are two problems to deal with. Firstly, due to the large character set, a small input set comparable to the English character set is needed for the end-to-end solution. Secondly, there are serious prosodic phrasing mistakes when the end-to-end method is applied to Chinese. In this paper, we will propose a solution for an end-to-end Chinese TTS system on the basis of Tacotron 2 and Wavenet vocoder. We will then add extra contextual information to improve the performance of prosodic phrasing. Our experiments have demonstrated the effectiveness of this proposal.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"26 1","pages":"7050-7054"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84378297","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 : 2019-05-12DOI: 10.1109/ICASSP.2019.8682603
B. I. Ahmad, P. Langdon, S. Godsill
Engineering Department, University of Cambridge, Trumpington Street, Cambridge, UK, CB2 1PZ In this paper, we introduce a generic Bayesian framework for inferring the intent of a tracked object, as early as possible, based on the available partial sensory observations. It treats the prediction problem, i.e. not estimating the object state such as position, within an object tracking formulation. This leads to a low-complexity implementation of the inference routine with minimal training requirements. The proposed approach utilises suitable stochastic, namely linear Gaussian, models to capture long term dependencies in the object trajectory as dictated by intent. Numerical examples are shown to demonstrate the efficacy of this framework.
{"title":"A Bayesian Framework for Intent Prediction in Object Tracking","authors":"B. I. Ahmad, P. Langdon, S. Godsill","doi":"10.1109/ICASSP.2019.8682603","DOIUrl":"https://doi.org/10.1109/ICASSP.2019.8682603","url":null,"abstract":"Engineering Department, University of Cambridge, Trumpington Street, Cambridge, UK, CB2 1PZ In this paper, we introduce a generic Bayesian framework for inferring the intent of a tracked object, as early as possible, based on the available partial sensory observations. It treats the prediction problem, i.e. not estimating the object state such as position, within an object tracking formulation. This leads to a low-complexity implementation of the inference routine with minimal training requirements. The proposed approach utilises suitable stochastic, namely linear Gaussian, models to capture long term dependencies in the object trajectory as dictated by intent. Numerical examples are shown to demonstrate the efficacy of this framework.","PeriodicalId":13203,"journal":{"name":"ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"89 1","pages":"8439-8443"},"PeriodicalIF":0.0,"publicationDate":"2019-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84394910","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：