This paper presents a speech recognition system that directly transcribes audio data with text, without requiring an intermediate phonetic representation. The system is based on a combination of the deep bidirectional LSTM recurrent neural network architecture and the Connectionist Temporal Classification objective function. A modification to the objective function is introduced that trains the network to minimise the expectation of an arbitrary transcription loss function. This allows a direct optimisation of the word error rate, even in the absence of a lexicon or language model. The system achieves a word error rate of 27.3% on the Wall Street Journal corpus with no prior linguistic information, 21.9% with only a lexicon of allowed words, and 8.2% with a trigram language model. Combining the network with a baseline system further reduces the error rate to 6.7%.

/pdf/towards-end-to-end-speech-recognition-with-recurrent-neural-58jrcs194u.pdf

Towards End-To-End Speech Recognition with Recurrent Neural Networks

This paper investigates data augmentation for deep neural network acoustic modeling based on label-preserving transformations to deal with data sparsity. Two data augmentation approaches, vocal tract length perturbation (VTLP) and stochastic feature mapping (SFM), are investigated for both deep neural networks (DNNs) and convolutional neural networks (CNNs). The approaches are focused on increasing speaker and speech variations of the limited training data such that the acoustic models trained with the augmented data are more robust to such variations. In addition, a two-stage data augmentation scheme based on a stacked architecture is proposed to combine VTLP and SFM as complementary approaches. Experiments are conducted on Assamese and Haitian Creole, two development languages of the IARPA Babel program, and improved performance on automatic speech recognition (ASR) and keyword search (KWS) is reported.

Data augmentation for deep neural network acoustic modeling

Augmenting datasets by transforming inputs in a way that does not change the label is a crucial ingredient of the state of the art methods for object recognition using neural networks. However this approach has (to our knowledge) not been exploited successfully in speech recognition (with or without neural networks). In this paper we lay the foundation for this approach, and show one way of augmenting speech datasets by transforming spectrograms, using a random linear warping along the frequency dimension. In practice this can be achieved by using warping techniques that are used for vocal tract length normalization (VTLN) - with the difference that a warp factor is generated randomly each time, during training, rather than tting a single warp factor to each training and test speaker (or utterance). At test time, a prediction is made by averaging the predictions over multiple warp factors. When this technique is applied to TIMIT using Deep Neural Networks (DNN) of dierent depths, the Phone Error Rate (PER) improved by an average of 0.65% on the test set. For a Convolutional neural network (CNN) with convolutional layer in the bottom, a gain of 1.0% was observed. These improvements were achieved without increasing the number of training epochs, and suggest that data transformations should be an important component of training neural networks for speech, especially for data limited projects.

/pdf/vocal-tract-length-perturbation-vtlp-improves-speech-41ndxjjol7.pdf

Vocal Tract Length Perturbation (VTLP) improves speech recognition

In recent times, deep artificial neural networks have achieved many successes in pattern recognition. Part of this success can be attributed to the reliance on big data to increase generalization. However, in the field of time series recognition, many datasets are often very small. One method of addressing this problem is through the use of data augmentation. In this paper, we survey data augmentation techniques for time series and their application to time series classification with neural networks. We propose a taxonomy and outline the four families in time series data augmentation, including transformation-based methods, pattern mixing, generative models, and decomposition methods. Furthermore, we empirically evaluate 12 time series data augmentation methods on 128 time series classification datasets with six different types of neural networks. Through the results, we are able to analyze the characteristics, advantages and disadvantages, and recommendations of each data augmentation method. This survey aims to help in the selection of time series data augmentation for neural network applications.

/pdf/an-empirical-survey-of-data-augmentation-for-time-series-5bdy19uq2z.pdf

An empirical survey of data augmentation for time series classification with neural networks.

A review of recent advances in visual speech decoding

In an effort to reduce the degradation in speech recognition performance caused by variation in vocal tract shape among speakers, a frequency warping approach to speaker normalization is investigated. A set of low complexity, maximum likelihood based frequency warping procedures have been applied to speaker normalization for a telephone based connected digit recognition task. This paper presents an efficient means for estimating a linear frequency warping factor and a simple mechanism for implementing frequency warping by modifying the filterbank in mel-frequency cepstrum feature analysis. An experimental study comparing these techniques to other well-known techniques for reducing variability is described. The results have shown that frequency warping is consistently able to reduce word error rate by 20% even for very short utterances.

/pdf/a-frequency-warping-approach-to-speaker-normalization-56s3hvnfo5.pdf

L. Lee

Papers

A frequency warping approach to speaker normalization