Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

"speech": models, code, and papers

Learning Similarity Functions for Pronunciation Variations

Jun 18, 2017
Einat Naaman, Yossi Adi, Joseph Keshet

A significant source of errors in Automatic Speech Recognition (ASR) systems is due to pronunciation variations which occur in spontaneous and conversational speech. Usually ASR systems use a finite lexicon that provides one or more pronunciations for each word. In this paper, we focus on learning a similarity function between two pronunciations. The pronunciations can be the canonical and the surface pronunciations of the same word or they can be two surface pronunciations of different words. This task generalizes problems such as lexical access (the problem of learning the mapping between words and their possible pronunciations), and defining word neighborhoods. It can also be used to dynamically increase the size of the pronunciation lexicon, or in predicting ASR errors. We propose two methods, which are based on recurrent neural networks, to learn the similarity function. The first is based on binary classification, and the second is based on learning the ranking of the pronunciations. We demonstrate the efficiency of our approach on the task of lexical access using a subset of the Switchboard conversational speech corpus. Results suggest that on this task our methods are superior to previous methods which are based on graphical Bayesian methods.

  Access Paper or Ask Questions

Learning Online Alignments with Continuous Rewards Policy Gradient

Aug 03, 2016
Yuping Luo, Chung-Cheng Chiu, Navdeep Jaitly, Ilya Sutskever

Sequence-to-sequence models with soft attention had significant success in machine translation, speech recognition, and question answering. Though capable and easy to use, they require that the entirety of the input sequence is available at the beginning of inference, an assumption that is not valid for instantaneous translation and speech recognition. To address this problem, we present a new method for solving sequence-to-sequence problems using hard online alignments instead of soft offline alignments. The online alignments model is able to start producing outputs without the need to first process the entire input sequence. A highly accurate online sequence-to-sequence model is useful because it can be used to build an accurate voice-based instantaneous translator. Our model uses hard binary stochastic decisions to select the timesteps at which outputs will be produced. The model is trained to produce these stochastic decisions using a standard policy gradient method. In our experiments, we show that this model achieves encouraging performance on TIMIT and Wall Street Journal (WSJ) speech recognition datasets.

  Access Paper or Ask Questions

A language model based approach towards large scale and lightweight language identification systems

Oct 13, 2015
Brij Mohan Lal Srivastava, Hari Krishna Vydana, Anil Kumar Vuppala, Manish Shrivastava

Multilingual spoken dialogue systems have gained prominence in the recent past necessitating the requirement for a front-end Language Identification (LID) system. Most of the existing LID systems rely on modeling the language discriminative information from low-level acoustic features. Due to the variabilities of speech (speaker and emotional variabilities, etc.), large-scale LID systems developed using low-level acoustic features suffer from a degradation in the performance. In this approach, we have attempted to model the higher level language discriminative phonotactic information for developing an LID system. In this paper, the input speech signal is tokenized to phone sequences by using a language independent phone recognizer. The language discriminative phonotactic information in the obtained phone sequences are modeled using statistical and recurrent neural network based language modeling approaches. As this approach, relies on higher level phonotactical information it is more robust to variabilities of speech. Proposed approach is computationally light weight, highly scalable and it can be used in complement with the existing LID systems.

* Under review at ICASSP 2016 

  Access Paper or Ask Questions

Denoising Deep Neural Networks Based Voice Activity Detection

Mar 04, 2013
Xiao-Lei Zhang, Ji Wu

Recently, the deep-belief-networks (DBN) based voice activity detection (VAD) has been proposed. It is powerful in fusing the advantages of multiple features, and achieves the state-of-the-art performance. However, the deep layers of the DBN-based VAD do not show an apparent superiority to the shallower layers. In this paper, we propose a denoising-deep-neural-network (DDNN) based VAD to address the aforementioned problem. Specifically, we pre-train a deep neural network in a special unsupervised denoising greedy layer-wise mode, and then fine-tune the whole network in a supervised way by the common back-propagation algorithm. In the pre-training phase, we take the noisy speech signals as the visible layer and try to extract a new feature that minimizes the reconstruction cross-entropy loss between the noisy speech signals and its corresponding clean speech signals. Experimental results show that the proposed DDNN-based VAD not only outperforms the DBN-based VAD but also shows an apparent performance improvement of the deep layers over shallower layers.

* This paper has been accepted by IEEE ICASSP-2013, and will be published online after May, 2013 

  Access Paper or Ask Questions

Efficient Arabic emotion recognition using deep neural networks

Oct 31, 2020
Ahmed Ali, Yasser Hifny

Emotion recognition from speech signal based on deep learning is an active research area. Convolutional neural networks (CNNs) may be the dominant method in this area. In this paper, we implement two neural architectures to address this problem. The first architecture is an attention-based CNN-LSTM-DNN model. In this novel architecture, the convolutional layers extract salient features and the bi-directional long short-term memory (BLSTM) layers handle the sequential phenomena of the speech signal. This is followed by an attention layer, which extracts a summary vector that is fed to the fully connected dense layer (DNN), which finally connects to a softmax output layer. The second architecture is based on a deep CNN model. The results on an Arabic speech emotion recognition task show that our innovative approach can lead to significant improvements (2.2% absolute improvements) over a strong deep CNN baseline system. On the other hand, the deep CNN models are significantly faster than the attention based CNN-LSTM-DNN models in training and classification.

  Access Paper or Ask Questions

Stochastic WaveNet: A Generative Latent Variable Model for Sequential Data

Jun 15, 2018
Guokun Lai, Bohan Li, Guoqing Zheng, Yiming Yang

How to model distribution of sequential data, including but not limited to speech and human motions, is an important ongoing research problem. It has been demonstrated that model capacity can be significantly enhanced by introducing stochastic latent variables in the hidden states of recurrent neural networks. Simultaneously, WaveNet, equipped with dilated convolutions, achieves astonishing empirical performance in natural speech generation task. In this paper, we combine the ideas from both stochastic latent variables and dilated convolutions, and propose a new architecture to model sequential data, termed as Stochastic WaveNet, where stochastic latent variables are injected into the WaveNet structure. We argue that Stochastic WaveNet enjoys powerful distribution modeling capacity and the advantage of parallel training from dilated convolutions. In order to efficiently infer the posterior distribution of the latent variables, a novel inference network structure is designed based on the characteristics of WaveNet architecture. State-of-the-art performances on benchmark datasets are obtained by Stochastic WaveNet on natural speech modeling and high quality human handwriting samples can be generated as well.

* ICML 2018 Workshop 

  Access Paper or Ask Questions

Integrating Categorical Features in End-to-End ASR

Oct 06, 2021
Rongqing Huang

All-neural, end-to-end ASR systems gained rapid interest from the speech recognition community. Such systems convert speech input to text units using a single trainable neural network model. E2E models require large amounts of paired speech text data that is expensive to obtain. The amount of data available varies across different languages and dialects. It is critical to make use of all these data so that both low resource languages and high resource languages can be improved. When we want to deploy an ASR system for a new application domain, the amount of domain specific training data is very limited. To be able to leverage data from existing domains is important for ASR accuracy in the new domain. In this paper, we treat all these aspects as categorical information in an ASR system, and propose a simple yet effective way to integrate categorical features into E2E model. We perform detailed analysis on various training strategies, and find that building a joint model that includes categorical features can be more accurate than multiple independently trained models.

* Submitted to ICASSP 2022 

  Access Paper or Ask Questions

Earnings-21: A Practical Benchmark for ASR in the Wild

Apr 28, 2021
Miguel Del Rio, Natalie Delworth, Ryan Westerman, Michelle Huang, Nishchal Bhandari, Joseph Palakapilly, Quinten McNamara, Joshua Dong, Piotr Zelasko, Miguel Jette

Commonly used speech corpora inadequately challenge academic and commercial ASR systems. In particular, speech corpora lack metadata needed for detailed analysis and WER measurement. In response, we present Earnings-21, a 39-hour corpus of earnings calls containing entity-dense speech from nine different financial sectors. This corpus is intended to benchmark ASR systems in the wild with special attention towards named entity recognition. We benchmark four commercial ASR models, two internal models built with open-source tools, and an open-source LibriSpeech model and discuss their differences in performance on Earnings-21. Using our recently released fstalign tool, we provide a candid analysis of each model's recognition capabilities under different partitions. Our analysis finds that ASR accuracy for certain NER categories is poor, presenting a significant impediment to transcript comprehension and usage. Earnings-21 bridges academic and commercial ASR system evaluation and enables further research on entity modeling and WER on real world audio.

* submitted to INTERSPEECH 2021 Update April 28th, 2021: We found and resolved an issue in our experimental evaluation that scored the LibriSpeech model at ~20% worse relative WER than the actual WER. The updated results do not affect our conclusions 

  Access Paper or Ask Questions

Sequence Transduction with Recurrent Neural Networks

Nov 14, 2012
Alex Graves

Many machine learning tasks can be expressed as the transformation---or \emph{transduction}---of input sequences into output sequences: speech recognition, machine translation, protein secondary structure prediction and text-to-speech to name but a few. One of the key challenges in sequence transduction is learning to represent both the input and output sequences in a way that is invariant to sequential distortions such as shrinking, stretching and translating. Recurrent neural networks (RNNs) are a powerful sequence learning architecture that has proven capable of learning such representations. However RNNs traditionally require a pre-defined alignment between the input and output sequences to perform transduction. This is a severe limitation since \emph{finding} the alignment is the most difficult aspect of many sequence transduction problems. Indeed, even determining the length of the output sequence is often challenging. This paper introduces an end-to-end, probabilistic sequence transduction system, based entirely on RNNs, that is in principle able to transform any input sequence into any finite, discrete output sequence. Experimental results for phoneme recognition are provided on the TIMIT speech corpus.

* First published in the International Conference of Machine Learning (ICML) 2012 Workshop on Representation Learning 

  Access Paper or Ask Questions