Acoustic Modeling for Google Home

Li, Bo; Sainath, Tara N.; Narayanan, Arun; Caroselli, Joe; Bacchiani, Michiel; Misra, Ananya; Shafran, Izhak; Sak, Haşim; Pundak, Golan; Chin, K. K.; Sim, Khe Chai; Weiss, Ron; Wilson, Kevin; Variani, Ehsan; Kim, Chanwoo; Siohan, Olivier; Weintraub, M.; McDermott, Erik; Rose, Richard C.; Shannon, Matt

doi:10.21437/interspeech.2017-234

Cited by 151 publications

(113 citation statements)

References 11 publications

Supporting

Mentioning

112

Contrasting

Unclassified

Order By: Relevance

“…The usefulness of these techniques, particularly for ASR, has been extensively studied, e.g., at the REVERB challenge [19] and the CHiME-3/4/5 challenges [20]- [22]. Moreover, advances in these techniques have led to recent progress on commercial devices, such as smart speakers [23]- [25].…”

Section: Introductionmentioning

confidence: 99%

Maximum likelihood convolutional beamformer for simultaneous denoising and dereverberation

Nakatani

Kinoshita

2019

2019 27th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

This article describes a probabilistic formulation of a Weighted Power minimization Distortionless response convolutional beamformer (WPD). The WPD unifies a weighted prediction error based dereverberation method (WPE) and a minimum power distortionless response beamformer (MPDR) into a single convolutional beamformer, and achieves simultaneous dereverberation and denoising in an optimal way. However, the optimization criterion is obtained simply by combining existing criteria without any clear theoretical justification. This article presents a generative model and a probabilistic formulation of a WPD, and derives an optimization algorithm based on a maximum likelihood estimation. We also describe a method for estimating the steering vector of the desired signal by utilizing WPE within the WPD framework to provide an effective and efficient beamformer for denoising and dereverberation.

show abstract

Section: Introductionmentioning

confidence: 99%

Maximum likelihood convolutional beamformer for simultaneous denoising and dereverberation

Nakatani

Kinoshita

2019

2019 27th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

show abstract

“…However it is still unclear how to design a suitable neural network architecture to exploit temporal and frequency information in deriving an effective speech emotional representation. In [14,15], 2D Time-Frequency (TF) LSTM and Grid-LSTM were proposed to model the variation over time and frequency for large scale automatic speech recognition (ASR). However, complex model architectures are prone to overfitting on a small scale dataset such as IEMOCAP [16].…”

Section: Introductionmentioning

confidence: 99%

An Attention Pooling Based Representation Learning Method for Speech Emotion Recognition

et al. 2018

View full text Add to dashboard Cite

This paper proposes an attention pooling based representation learning method for speech emotion recognition (SER). The emotional representation is learned in an end-to-end fashion by applying a deep convolutional neural network (CNN) directly to spectrograms extracted from speech utterances. Motivated by the success of GoogleNet, two groups of filters with different shapes are designed to capture both temporal and frequency domain context information from the input spectrogram. The learned features are concatenated and fed into the subsequent convolutional layers. To learn the final emotional representation, a novel attention pooling method is further proposed. Compared with the existing pooling methods, such as max-pooling and average-pooling, the proposed attention pooling can effectively incorporate class-agnostic bottom-up, and class-specific top-down, attention maps. We conduct extensive evaluations on benchmark IEMOCAP data to assess the effectiveness of the proposed representation. Results demonstrate a recognition performance of 71.8% weighted accuracy (WA) and 68% unweighted accuracy (UA) over four emotions, which outperforms the state-of-the-art method by about 3% absolute for WA and 4% for UA.

show abstract

“…This improvement has come about from the shift from Gaussian Mixture Model (GMM) to the Feed-Forward Deep Neural Networks (FF-DNNs), FF-DNNs to Recurrent Neural Network (RNN) and in particular the Long Short-Term Memory (LSTM) networks [9]. Thanks to these advances, voice assistant devices such as Google Home [2,10] , Amazon Alexa or Samsung Bixby [11] are being used at many homes and on personal devices.…”

Section: Introductionmentioning

confidence: 99%

End-to-End Training of a Large Vocabulary End-to-End Speech Recognition System

Kim

Shin

Singh

et al. 2019

2019 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)

Self Cite

View full text Add to dashboard Cite

In this paper, we present an end-to-end training framework for building state-of-the-art end-to-end speech recognition systems.Our training system utilizes a cluster of Central Processing Units (CPUs) and Graphics Processing Units (GPUs). The entire data reading, large scale data augmentation, neural network parameter updates are all performed "on-the-fly". We use vocal tract length perturbation [1] and an acoustic simulator [2] for data augmentation. The processed features and labels are sent to the GPU cluster. The Horovod allreduce approach is employed to train neural network parameters. We evaluated the effectiveness of our system on the standard Librispeech corpus [3] and the 10,000-hr anonymized Bixby English dataset. Our end-to-end speech recognition system built using this training infrastructure showed a 2.44 % WER on test-clean of the LibriSpeech test set after applying shallow fusion with a Transformer language model (LM). For the proprietary English Bixby open domain test set, we obtained a WER of 7.92 % using a Bidirectional Full Attention (BFA) end-to-end model after applying shallow fusion with an RNN-LM. When the monotonic chunckwise attention (MoCha) based approach is employed for streaming speech recognition, we obtained a WER of 9.95 % on the same Bixby open domain test set.

show abstract

Acoustic Modeling for Google Home

Cited by 151 publications

References 11 publications

Maximum likelihood convolutional beamformer for simultaneous denoising and dereverberation

Maximum likelihood convolutional beamformer for simultaneous denoising and dereverberation

An Attention Pooling Based Representation Learning Method for Speech Emotion Recognition

End-to-End Training of a Large Vocabulary End-to-End Speech Recognition System

Contact Info

Product

Resources

About