Attentive batch normalization for lstm-based acoustic modeling of speech recognition
National Engineering Laboratory for Speech and Language Information Processing（国家语音与语言信息处理工程实验室）；University of Science and Technology of China（中国科学技术大学）
Batch normalization (BN) is an effective method to accelerate model training and improve the generalization performance of neural networks. In this paper, we propose an improved batch normalization technique called attentive batch normalization (ABN) in Long Short Term Memory (LSTM) based acoustic modeling for automatic speech recognition (ASR). In the proposed method, an auxiliary network is used to dynamically generate the scaling and shifting parameters in batch normalization, and attention mechanisms are introduced to improve their regularized performance. Furthermore, two schemes, frame-level and utterance-level ABN, are investigated. We evaluate our proposed methods on Mandarin and Uyghur ASR tasks, respectively. The experimental results show that the proposed ABN greatly improves the performance of batch normalization in terms of transcription accuracy for both languages.
The effective exploitation of richer contextual information in language models (LMs) is a long-standing research problem for automatic speech recognition (ASR). A cross-utterance LM (CULM) is proposed in this paper, which augments the input to a standard long short-term memory (LSTM) LM with a context vector derived from past and future utterances using an extraction network. The extraction network uses another LSTM to encode surrounding utterances into vectors which are integrated into a context vector using either a projection of LSTM final hidden states, or a multi-head self-attentive layer…
We present a comprehensive study of deep bidirectional long short-term memory (LSTM) recurrent neural network (RNN) based acoustic models for automatic speech recognition (ASR). We study the effect of size and depth and train models of up to 8 layers. We investigate the training aspect and study different variants of optimization methods, batching, truncated backpropagation, different regularization techniques such as dropout and $L_2$ regularization, and different gradient clipping variants. The major part of the experimental analysis was performed on the Quaero corpus. Additional experiments…
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；University of Chinese Academy of Sciences（中国科学院大学）
LSTM-based speaker verification usually uses a fixed-length local segment randomly truncated from an utterance to learn the utterance-level speaker embedding, while using the average embedding of all segments of a test utterance to verify the speaker, which results in a critical mismatch between testing and training. This mismatch degrades the performance of speaker verification, especially when the durations of training and testing utterances are very different. To alleviate this issue, we propose the deep segment attentive embedding method to learn the unified speaker embeddings for utteranc…
Acoustic echo cancellation (AEC) plays an important role in the full-duplex speech communication as well as the front-end speech enhancement for recognition in the conditions when the loudspeaker plays back. In this paper, we present an all-deep-learning framework that implicitly estimates the second order statistics of echo/noise and target speech, and jointly solves echo and noise suppression through an attention based recurrent neural network. The proposed model outperforms the state-of-the-art joint echo cancellation and speech enhancement method F-T-LSTM in terms of objective speech quali…
In this paper, we introduce a novel language identification system based on conformer layers. We propose an attentive temporal pooling mechanism to allow the model to carry information in long-form audio via a recurrent form, such that the inference can be performed in a streaming fashion. Additionally, we investigate two domain adaptation approaches to allow adapting an existing language identification model without retraining the model parameters for a new domain. We perform a comparative study of different model topologies under different constraints of model size, and find that conformer-b…
Humelo Inc.（Humelo公司）；Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院（KAIST））；KAIST Institute for Artificial Intelligence（KAIST人工智能研究所）
Generalized end-to-end (GE2E) model is widely used in speaker verification (SV) fields due to its expandability and generality regardless of specific languages. However, the long-short term memory (LSTM) based on GE2E has two limitations: First, the embedding of GE2E suffers from vanishing gradient, which leads to performance degradation for very long input sequences. Secondly, utterances are not represented as a properly fixed dimensional vector. In this paper, to overcome issues mentioned above, we propose a novel framework for SV, end-to-end trainable self-attentive shallow network (SASN), …
We investigate the use of generative adversarial networks (GANs) in speech dereverberation for robust speech recognition. GANs have been recently studied for speech enhancement to remove additive noises, but there still lacks of a work to examine their ability in speech dereverberation and the advantages of using GANs have not been fully established. In this paper, we provide deep investigations in the use of GAN-based dereverberation front-end in ASR. First, we study the effectiveness of different dereverberation networks (the generator in GAN) and find that LSTM leads a significant improveme…
In this paper, we propose and investigate a variety of distributed deep learning strategies for automatic speech recognition (ASR) and evaluate them with a state-of-the-art Long short-term memory (LSTM) acoustic model on the 2000-hour Switchboard (SWB2000), which is one of the most widely used datasets for ASR performance benchmark. We first investigate what are the proper hyper-parameters (e.g., learning rate) to enable the training with sufficiently large batch size without impairing the model accuracy. We then implement various distributed strategies, including Synchronous (SYNC), Asynchron…
There is a growing interest in the speech community in developing Recurrent Neural Network Transducer (RNN-T) models for automatic speech recognition (ASR) applications. RNN-T is trained with a loss function that does not enforce temporal alignment of the training transcripts and audio. As a result, RNN-T models built with uni-directional long short term memory (LSTM) encoders tend to wait for longer spans of input audio, before streaming already decoded ASR tokens. In this work, we propose a modification to the RNN-T loss function and develop Alignment Restricted RNN-T (Ar-RNN-T) models, whic…
One of the most important parts of an end-to-end speaker verification system is the speaker embedding generation. In our previous paper, we reported that shortcut connections-based multi-layer aggregation improves the representational power of the speaker embedding. However, the number of model parameters is relatively large and the unspecified variations increase in the multi-layer aggregation. Therefore, we propose a self-attentive multi-layer aggregation with feature recalibration and normalization for end-to-end speaker verification system. To reduce the number of model parameters, the Res…
Recently sequence-to-sequence models have started to achieve state-of-the-art performance on standard speech recognition tasks when processing audio data in batch mode, i.e., the complete audio data is available when starting processing. However, when it comes to performing run-on recognition on an input stream of audio data while producing recognition results in real-time and with low word-based latency, these models face several challenges. For many techniques, the whole audio sequence to be decoded needs to be available at the start of the processing, e.g., for the attention mechanism or th…
Spectrogram is commonly used as the input feature of deep neural networks to learn the high(er)-level time-frequency pattern of speech signal for speech emotion recognition (SER). \textcolor{black}{Generally, different emotions correspond to specific energy activations both within frequency bands and time frames on spectrogram, which indicates the frequency and time domains are both essential to represent the emotion for SER. However, recent spectrogram-based works mainly focus on modeling the long-term dependency in time domain, leading to these methods encountering the following two issues: …
In this work we design a neural network for recognizing emotions in speech, using the IEMOCAP dataset. Following the latest advances in audio analysis, we use an architecture involving both convolutional layers, for extracting high-level features from raw spectrograms, and recurrent ones for aggregating long-term dependencies. We examine the techniques of data augmentation with vocal track length perturbation, layer-wise optimizer adjustment, batch normalization of recurrent layers and obtain highly competitive results of 64.5% for weighted accuracy and 61.7% for unweighted accuracy on four em…
In this paper, we propose an ensemble of deep neural networks along with data augmentation (DA) learned using effective speech-based features to recognize emotions from speech. Our ensemble model is built on three deep neural network-based models. These neural networks are built using the basic local feature acquiring blocks (LFAB) which are consecutive layers of dilated 1D Convolutional Neural networks followed by the max pooling and batch normalization layers. To acquire the long-term dependencies in speech signals further two variants are proposed by adding Gated Recurrent Unit (GRU) and Lo…
The use of future contextual information is typically shown to be helpful for acoustic modeling. Recently, we proposed a RNN model called minimal gated recurrent unit with input projection (mGRUIP), in which a context module namely temporal convolution, is specifically designed to model the future context. This model, mGRUIP with context module (mGRUIP-Ctx), has been shown to be able of utilizing the future context effectively, meanwhile with quite low model latency and computation cost. In this paper, we continue to improve mGRUIP-Ctx with two revisions: applying BN methods and enlarging mode…
Traditionally, music was treated as an analogue signal and was generated manually. In recent years, music is conspicuous to technology which can generate a suite of music automatically without any human intervention. To accomplish this task, we need to overcome some technical challenges which are discussed descriptively in this paper. A brief introduction about music and its components is provided in the paper along with the citation and analysis of related work accomplished by different authors in this domain. Main objective of this paper is to propose an algorithm which can be used to genera…
RNN-T models have gained popularity in the literature and in commercial systems because of their competitiveness and capability of operating in online streaming mode. In this work, we conduct an extensive study comparing several prediction network architectures for both monotonic and original RNN-T models. We compare 4 types of prediction networks based on a common state-of-the-art Conformer encoder and report results obtained on Librispeech and an internal medical conversation data set. Our study covers both offline batch-mode and online streaming scenarios. In contrast to some previous works…
Budapest University of Technology and Economics（布达佩斯技术与经济大学）；Department of Telecommunications and Media Informatics（电信与媒体信息学系）；Faculty of Electrical Engineering and Informatics（电气工程与信息学院）；Department of Telecommunications（电信系）；Media Informatics Faculty of Electrical Engineering（电气工程媒体信息学院）；Informatics Budapest University of Technology（布达佩斯技术大学信息学院）
MOS (Mean Opinion Score) is a subjective method used for the evaluation of a system's quality. Telecommunications (for voice and video), and speech synthesis systems (for generated speech) are a few of the many applications of the method. While MOS tests are widely accepted, they are time-consuming and costly since human input is required. In addition, since the systems and subjects of the tests differ, the results are not really comparable. On the other hand, a large number of previous tests allow us to train machine learning models that are capable of predicting MOS value. By automatically p…
