Relative Word Error Rate Reduction Research Articles

DNN-based Multilingual Acoustic Modeling for Four Ethiopian Languages

In this paper, we present the results of experiments conducted on multilingual acoustic modeling in the development of an Automatic Speech Recognition (ASR) system using speech data of phonetically much related Ethiopian languages (Amharic, Tigrigna, Oromo and Wolaytta) with multilingual (ML) mix and multitask approaches. The use of speech data from only phonetically much related languages brought improvement over results reported in a previous work that used 26 languages (including the four languages). A maximum Word Error Rate (WER) reduction from 25.03% (in the previous work) to 21.52% has been achieved for Wolaytta, which is a relative WER reduction of 14.02%. As a result of using multilingual acoustic modeling for the development of an automatic speech recognition (ASR) system, a relative WER reduction of up to 7.36% (a WER reduction from 23.23% to 21.52%) has been achieved over a monolingual ASR. Compared to the ML mix, the multitask approach brought a better performance improvement (a relative WERs reduction of up to 5.9%). Experiments have also been conducted using Amharic and Tigrigna in a pair and Oromo and Wolaytta in another pair. The results of the experiments showed that languages with a relatively better language resources for lexical and language modeling (Amharic and Tigrigna) benefited from the use of speech data from only two languages. Generally, the findings show that the use of speech corpora of phonetically related languages with the multitask multilingual modeling approach for the development of ASR systems for less-resourced languages is a promising solution.

Language fusion via adapters for low-resource speech recognition

Data scarcity makes low-resource speech recognition systems suffer from severe overfitting. Although fine-tuning addresses this issue to some extent, it leads to parameter-inefficient training. In this paper, a novel language knowledge fusion method, named LanFusion, is proposed. It is built on the recent popular adapter-tuning technique, thus maintaining better parameter efficiency compared with conventional fine-tuning methods. LanFusion is a two-stage method. Specifically, multiple adapters are first trained on several source languages to extract language-specific and language-invariant knowledge. Then, the trained adapters are re-trained on the target low-resource language to fuse the learned knowledge. Compared with Vanilla-adapter, LanFusion obtains a relative average word error rate (WER) reduction of 9.8% and 8.6% on the Common Voice and FLEURS corpora, respectively. Extensive experiments demonstrate the proposed method is not only simple and effective but also parameter-efficient. Besides, using source languages that are geographically similar to the target language yields better results on both datasets.

Adapting Pre-Trained Self-Supervised Learning Model for Speech Recognition with Light-Weight Adapters

Self-supervised learning (SSL) is an effective way of learning rich and transferable speech representations from unlabeled data to benefit downstream tasks. However, effectively incorporating a pre-trained SSL model into an automatic speech recognition (ASR) system remains challenging. In this paper, we propose a network architecture with light-weight adapters to adapt a pre-trained SSL model for an end-to-end (E2E) ASR. An adapter is introduced in each SSL network layer and trained on the downstream ASR task, while the parameters of the pre-trained SSL network layers remain unchanged. By carrying over all pre-trained parameters, we avoid the catastrophic forgetting problem. At the same time, we allow the network to quickly adapt to ASR task with light-weight adapters. The experiments using LibriSpeech and Wall Street Journal (WSJ) datasets show that (1) the proposed adapter-based fine-tuning consistently outperforms full-fledged training in low-resource scenarios, with up to 17.5%/12.2% relative word error rate (WER) reduction on the 10 min LibriSpeech split; (2) the adapter-based adaptation also shows competitive performance in high-resource scenarios, which further validates the effectiveness of the adapters.

Semisupervised Speech Data Extraction from Basque Parliament Sessions and Validation on Fully Bilingual Basque–Spanish ASR

In this paper, a semisupervised speech data extraction method is presented and applied to create a new dataset designed for the development of fully bilingual Automatic Speech Recognition (ASR) systems for Basque and Spanish. The dataset is drawn from an extensive collection of Basque Parliament plenary sessions containing frequent code switchings. Since session minutes are not exact, only the most reliable speech segments are kept for training. To that end, we use phonetic similarity scores between nominal and recognized phone sequences. The process starts with baseline acoustic models trained on generic out-of-domain data, then iteratively updates the models with the extracted data and applies the updated models to refine the training dataset until the observed improvement between two iterations becomes small enough. A development dataset, involving five plenary sessions not used for training, has been manually audited for tuning and evaluation purposes. Cross-validation experiments (with 20 random partitions) have been carried out on the development dataset, using the baseline and the iteratively updated models. On average, Word Error Rate (WER) reduces from 16.57% (baseline) to 4.41% (first iteration) and further to 4.02% (second iteration), which corresponds to relative WER reductions of 73.4% and 8.8%, respectively. When considering only Basque segments, WER reduces on average from 16.57% (baseline) to 5.51% (first iteration) and further to 5.13% (second iteration), which corresponds to relative WER reductions of 66.7% and 6.9%, respectively. As a result of this work, a new bilingual Basque–Spanish resource has been produced based on Basque Parliament sessions, including 998 h of training data (audio segments + transcriptions), a development set (17 h long) designed for tuning and evaluation under a cross-validation scheme and a fully bilingual trigram language model.

Wav2vec‐MoE: An unsupervised pre‐training and adaptation method for multi‐accent ASR

AbstractIn real life, either the subjective factors of speakers or the objective environment degrades the performance of automatic speech recognition (ASR). This study focuses on one of the subjective factors, accented speech, and attempts to build a multi‐accent ASR system to solve the degradation caused by different accents, one of whose characteristic is the low resource. To deal with the challenge of the low‐resource data and the different speech styles, a wav2vec‐MoE (mixture of experts) is proposed to adapt the wav2vec 2.0 for multi‐accent ASR. In the wav2vec‐MoE, a domain MoE is developed by introducing pseudo‐domain information in the pre‐training stage, where the domain denotes a collection of speech varied by the same influence factors. The MoE is trained with two strategies according to the proposed domain mismatch assessment between unlabeled speech and target speech, without requiring any explicit domain information. Experiments show that the wav2vec‐MoE achieves a 14.69% relative word error rate reduction (WERR) on the AESRC2020 accent dataset and an 8.79% relative WERR on the Common Voice English dataset.

Combining hybrid DNN-HMM ASR systems with attention-based models using lattice rescoring

The traditional hybrid deep neural network (DNN)–hidden Markov model (HMM) system and attention-based encoder–decoder (AED) model are both commonly used automatic speech recognition (ASR) approaches with distinct characteristics and advantages. While hybrid systems are per-frame-based and highly modularised to leverage external phonetic and linguistic knowledge, AED models operate on a per-label basis and jointly learn the acoustic and language information using a single model in an end-to-end trainable fashion. In this paper, we propose combining these two approaches in a two-pass rescoring framework. The first-pass uses hybrid ASR systems to facilitate streaming and controllable ASR, and the second-pass re-scores the N-best hypotheses or lattices produced by the first-pass hybrid DNN-HMM system with AED models. We also propose an improved algorithm for lattice rescoring with AED models. Experiments show the combined two-pass systems achieve competitive performance without using extra speech or text data on two standard ASR tasks. For the 80-hour AMI IHM dataset, the combined system has a 13.7% word error rate (WER) on the evaluation set and is up to a 29% relative WER reduction over the individual systems. For the 300-hour Switchboard dataset, the WERs of the combined system are 5.7% and 12.1% on Switchboard and CallHome subsets of Hub5’00, and 13.2% and 7.6% on Switchboard Cellular and Fisher subsets of RT03, and are up to a 33% relative reduction in WER over the individual systems.

Morphology aware data augmentation with neural language models for online hybrid ASR

AbstractRecognition of Hungarian conversational telephone speech is challenging due to the informal style and morphological richness of the language. Neural Network Language Models (NNLMs) can provide remedy for the high perplexity of the task; however, their high complexity makes them very difficult to apply in the first (single) pass of an online system. Recent studies showed that a considerable part of the knowledge of NNLMs can be transferred to traditional n-grams by using neural text generation based data augmentation. Data augmentation with NNLMs works well for isolating languages; however, we show that it causes a vocabulary explosion in a morphologically rich language. Therefore, we propose a new, morphology aware neural text augmentation method, where we retokenize the generated text into statistically derived subwords. We compare the performance of word-based and subword-based data augmentation techniques with recurrent and Transformer language models and show that subword-based methods can significantly improve the Word Error Rate (WER) while greatly reducing vocabulary size and memory requirements. Combining subword-based modeling and neural language model-based data augmentation, we were able to achieve 11% relative WER reduction and preserve real-time operation of our conversational telephone speech recognition system. Finally, we also demonstrate that subword-based neural text augmentation outperforms the word-based approach not only in terms of overall WER but also in recognition of Out-of-Vocabulary (OOV) words.

Data Augmentation Using Spectral Warping for Low Resource Children ASR

In low resource children automatic speech recognition (ASR) the performance is degraded due to limited acoustic and speaker variability available in small datasets. In this paper, we propose a spectral warping based data augmentation method to capture more acoustic and speaker variability. This is carried out by warping the linear prediction (LP) spectra computed from speech data. The warped LP spectra computed in a frame-based manner are used with the corresponding LP residuals to synthesize speech to capture more variability. The proposed augmentation method is shown to improve the ASR system performance over the baseline system. We have compared the proposed method with four well-known data augmentation methods: pitch scaling, speaking rate, SpecAug and vocal tract length perturbation (VTLP), and found that the proposed method performs the best. Further, we have combined the proposed method with these existing data augmentation methods to improve the ASR system performance even more. The combined system consisting of the original data, VTLP, SpecAug and the proposed spectral warping method gave the best performance by a relative word error rate reduction of 32.13% and 10.51% over the baseline system for Punjabi children and TLT-school corpus, respectively. The proposed spectral warping method is publicly available at https://github.com/kathania/Spectral-Warping.

Reliability-Based Large-Vocabulary Audio-Visual Speech Recognition.

Audio-visual speech recognition (AVSR) can significantly improve performance over audio-only recognition for small or medium vocabularies. However, current AVSR, whether hybrid or end-to-end (E2E), still does not appear to make optimal use of this secondary information stream as the performance is still clearly diminished in noisy conditions for large-vocabulary systems. We, therefore, propose a new fusion architecture—the decision fusion net (DFN). A broad range of time-variant reliability measures are used as an auxiliary input to improve performance. The DFN is used in both hybrid and E2E models. Our experiments on two large-vocabulary datasets, the Lip Reading Sentences 2 and 3 (LRS2 and LRS3) corpora, show highly significant improvements in performance over previous AVSR systems for large-vocabulary datasets. The hybrid model with the proposed DFN integration component even outperforms oracle dynamic stream-weighting, which is considered to be the theoretical upper bound for conventional dynamic stream-weighting approaches. Compared to the hybrid audio-only model, the proposed DFN achieves a relative word-error-rate reduction of 51% on average, while the E2E-DFN model, with its more competitive audio-only baseline system, achieves a relative word error rate reduction of 43%, both showing the efficacy of our proposed fusion architecture.

Multilingual speech recognition for GlobalPhone languages

In this paper, we present our investigations towards the development of multilingual automatic speech recognition (ML ASR) systems using the GlobalPhone database. In addition to GlobalPhone, we have included 4 Ethiopian languages (Amharic, Oromo, Tigrigna and Wolaytta), as well as Uyghur and English in our investigation. In order to see the impact of language relatedness in ML ASR training, we have analyzed both phonetic overlap and morphological complexity of the languages. Deep Neural Network based ML ASR systems have been developed using ML mix, transfer and multitask learning approaches. Relative word error rate (WER) reductions up to 33.21% have been achieved as a result of using resources of other languages in ML acoustic model training. Our experimental results show that languages with small amounts of monolingual training data benefit a lot from ML training. Moreover, using phonetically related languages in ML training is more beneficiary than using phonetically less related languages. It seems that the nature of the corpus (single or mixed domain, noisy or noise free, etc.) has also an impact in ML training although it is not as important as the phonetic relatedness of the languages.

Combining Frame-Synchronous and Label-Synchronous Systems for Speech Recognition

Commonly used automatic speech recognition (ASR) systems can be classified into frame-synchronous and label-synchronous categories, based on whether the speech is decoded on a per-frame or per-label basis. Frame-synchronous systems, such as traditional hidden Markov model systems, can easily incorporate existing knowledge and can support streaming ASR applications. Label-synchronous systems, based on attention-based encoder-decoder models, can jointly learn the acoustic and language information with a single model, which can be regarded as audio-grounded language models. In this paper, we propose rescoring the N-best hypotheses or lattices produced by a first-pass frame-synchronous system with a label-synchronous system in a second-pass. By exploiting the complementary modelling of the different approaches, the combined two-pass systems achieve competitive performance without using any extra speech or text data on two standard ASR tasks. For the 80-hour AMI IHM dataset, the combined system has a 13.7% word error rate (WER) on the evaluation set, which is up to a 29% relative WER reduction over the individual systems. For the 300-hour Switchboard dataset, the WERs of the combined system are 5.7% and 12.1% on Switchboard and CallHome subsets of Hub5'00, and 13.2% and 7.6% on Switchboard Cellular and Fisher subsets of RT03, up to a 33% relative reduction in WER over the individual systems.

Optimizing Data Usage for Low-Resource Speech Recognition

Automatic speech recognition has made huge progress recently. However, the current modeling strategy still suffers a large performance degradation when facing the low-resource languages with limited training data. In this paper, we propose a series of methods to optimize the data usage for low-resource speech recognition. Multilingual speech recognition helps a lot in low-resource scenarios. The correlation and similarity between languages are further exploited for multilingual pretraining in our work. We utilize the posterior of the target language extracted from a language classifier to perform data weighing on training samples, which assists the model in being more biased towards the target language during pretraining. Furthermore, dynamic curriculum learning for data allocation and length perturbation for data augmentation are also designed. All these three methods form the new strategy on optimized data usage for low-resource languages. We evaluate the proposed method using rich resource languages for pretraining (PT) and finetuning (FT) the model on the target language with limited data. Experimental results show that the proposed data usage method obtains a 15 to 25&#x0025; relative word error rate reduction for different target languages compared with the commonly adopted multilingual PT+FT method on CommonVoice dataset. The same improvement and conclusion are also observed on Babel dataset with conversational telephone speech, and <inline-formula><tex-math notation="LaTeX">$\sim$</tex-math></inline-formula>40&#x0025; relative character error rate reduction can be obtained for the target low-resource language.

End-to-End Dereverberation, Beamforming, and Speech Recognition in a Cocktail Party

Far-field multi-speaker automatic speech recognition (ASR) has drawn increasing attention in recent years. Most existing methods feature a signal processing frontend and an ASR backend. In realistic scenarios, these modules are usually trained separately or progressively, which suffers from either inter-module mismatch or a complicated training process. In this paper, we propose an end-to-end multi-channel model that jointly optimizes the speech enhancement (including speech dereverberation, denoising, and separation) frontend and the ASR backend as a single system. To the best of our knowledge, this is the first work that proposes to optimize dereverberation, beamforming, and multi-speaker ASR in a fully end-to-end manner. The frontend module consists of a weighted prediction error (WPE) based submodule for dereverberation and a neural beamformer for denoising and speech separation. For the backend, we adopt a widely used end-to-end (E2E) ASR architecture. It is worth noting that the entire model is differentiable and can be optimized in a fully end-to-end manner using only the ASR criterion, without the need of parallel signal-level labels. We evaluate the proposed model on several multi-speaker benchmark datasets, and experimental results show that the fully E2E ASR model can achieve competitive performance on both noisy and reverberant conditions, with over 30% relative word error rate (WER) reduction over the single-channel baseline systems.

Towards Contextual Spelling Correction for Customization of End-to-End Speech Recognition Systems

Contextual biasing is an important and challenging task for end-to-end automatic speech recognition (ASR) systems, which aims to achieve better recognition performance by biasing the ASR system to particular context phrases such as person names, music list, proper nouns, etc. Existing methods mainly include contextual LM biasing and adding bias encoder into end-to-end ASR models. In this work, we introduce a novel approach to do contextual biasing by adding a contextual spelling correction model on top of the end-to-end ASR system. We incorporate contextual information into a sequence-to-sequence spelling correction model with a shared context encoder. The proposed model includes two different mechanisms: autoregressive (AR) and non-autoregressive (NAR). We also propose filtering algorithms to handle large-size context lists, and performance balancing mechanisms to control the biasing degree of the model. The proposed model is a general biasing solution which is domain-insensitive and can be adopted in different scenarios. Experiments show that the proposed method achieves as much as 51% relative word error rate (WER) reduction over ASR system and outperforms traditional biasing methods. Compared to the AR solution, the NAR model reduces model size by 43.2% and speeds up inference by 2.1 times.

Auxiliary Networks for Joint Speaker Adaptation and Speaker Change Detection

Speaker adaptation and speaker change detection have both been studied extensively to improve automatic speech recognition (ASR). In many cases, these two problems are investigated separately: speaker change detection is implemented first to obtain single-speaker regions, and speaker adaptation is then performed using the derived speaker segments for improved ASR. However, in an online setting, we want to achieve both goals in a single pass. In this study, we propose a neural network architecture that learns a speaker embedding from which it can perform both speaker adaptation for ASR and speaker change detection. The proposed speaker embedding is computed using self-attention based on an auxiliary network attached to a main ASR network. ASR adaptation is then performed by subtracting, from the main network activations, a segment dependent affine transformation of the learned speaker embedding. In experiments on a broadcast news dataset and the Switchboard conversational dataset, we test our system on utterances with a change point in them and show that the proposed method achieves significantly better performance as compared to the unadapted main network (10–14% relative reduction in word error rate (WER)). The proposed architecture also outperforms three different speaker segmentation methods followed by ASR (around 10% relative reduction in WER).

Online Speaker Adaptation Using Memory-Aware Networks for Speech Recognition

In our previous work, we introduced our attention-based speaker adaptation method, which has been proved to be an efficient online speaker adaptation method for real-time speech recognition. In this paper, we present a more complete framework of this method named memory-aware networks, which consists of the main network, the memory module, the attention module and the connection module. A gate mechanism and a multiple-connections strategy are presented to connect the memory with the main network in order to take full advantage of the memory. An auxiliary speaker classification task is provided to improve the accuracy of the attention module. The fixed-size ordinally forgetting encoding method is used together with average pooling to gather both short-term and long-term information. Furthermore, instead of only using traditional speaker embeddings such as i-vectors or d-vectors as the memory, we design a new form of memory called residual vectors, which can represent different pronunciation habits. Experiments on both the Switchboard and AISHELL-2 tasks show that our method can perform online speaker adaptation very well with no additional adaptation data and with only a relative 3% increase in decoding computation complexity. Under the cross-entropy criterion, our method achieves a relative word error rate reduction of 9.4% and 8.3% compared to that of the speaker-independent model on the Switchboard task and the AISHELL-2 task, respectively, and approximately 7.0% compared to that of the traditional d-vector-based speaker adaptation method.

Improving Knowledge Distillation of CTC-Trained Acoustic Models With Alignment-Consistent Ensemble and Target Delay

Knowledge distillation (KD) has been widely used to improve the performance of a simpler student model by imitating the outputs or intermediate representations of a more complex teacher model. The most commonly used KD technique is to minimize a Kullback-Leibler divergence between the output distributions of the teacher and student models. When it is applied to compressing acoustic models trained with a connectionist temporal classification (CTC) criterion, an assumption is made that the teacher and student share the same frame-level feature-transcription alignment. However, frame-level alignments learned by teachers can be inaccurate and unstable due to the lack of fine-grained frame-level guidance during CTC training. Forcing student to learn inaccurate alignments will lead to limited performance improvements. In this article, we investigate building powerful teacher models with more accurate and stable feature-transcription alignments. We achieve this goal by using a novel alignment-consistent ensemble (ACE) technique, where all models within an ensemble are jointly trained along with a regularization term to encourage consistent and stable alignments. With well-trained deep bidirectional LSTM (DBLSTM) ACE as a teacher, we can directly use the traditional frame-wise KD method to train DBLSTM students. When applying KD to transfer knowledge from a DBLSTM ACE to a deep unidirectional LSTM (DLSTM) student, a simple yet effective target delay technique is proposed to handle the alignment difference between bidirectional and unidirectional models. Experimental results on Switchboard-I speech recognition task show that, with DBLSTM ACE as a teacher, the simple frame-wise KD method can achieve competitive or better performance than other complex KD methods on DBLSTM students. When applying KD to build DLSTM students from DBLSTM teachers, our proposed target delay technique can achieve relative word error rate reductions of 14.2% $\sim$ 14.8% compared with the models trained from scratch, which outperforms other carefully-designed KD methods.

Multi-Stream End-to-End Speech Recognition

Attention-based methods and Connectionist Temporal Classification (CTC) network have been promising research directions for end-to-end (E2E) Automatic Speech Recognition (ASR). The joint CTC/Attention model has achieved great success by utilizing both architectures during multi-task training and joint decoding. In this article, we present a multi-stream framework based on joint CTC/Attention E2E ASR with parallel streams represented by separate encoders aiming to capture diverse information. On top of the regular attention networks, the Hierarchical Attention Network (HAN) is introduced to steer the decoder toward the most informative encoders. A separate CTC network is assigned to each stream to force monotonic alignments. Two representative framework have been proposed and discussed, which are Multi-Encoder Multi-Resolution (MEM-Res) framework and Multi-Encoder Multi-Array (MEM-Array) framework, respectively. In MEM-Res framework, two heterogeneous encoders with different architectures, temporal resolutions and separate CTC networks work in parallel to extract complementary information from same acoustics. Experiments are conducted on Wall Street Journal (WSJ) and CHiME-4, resulting in relative Word Error Rate (WER) reduction of $\text{18.0}\!-\!\text{32.1}\%$ and the best WER of $\text{3.6}\%$ in the WSJ eval92 test set. The MEM-Array framework aims at improving the far-field ASR robustness using multiple microphone arrays which are activated by separate encoders. Compared with the best single-array results, the proposed framework has achieved relative WER reduction of $\text{3.7}\%$ and $\text{9.7}\%$ in AMI and DIRHA multi-array corpora, respectively, which also outperforms conventional fusion strategies.

Speech Enhancement Based on Teacher–Student Deep Learning Using Improved Speech Presence Probability for Noise-Robust Speech Recognition

In this paper, we propose a novel teacher-student learning framework for the preprocessing of a speech recognizer, leveraging the online noise tracking capabilities of improved minima controlled recursive averaging (IMCRA) and deep learning of nonlinear interactions between speech and noise. First, a teacher model with deep architectures is built to learn the target of ideal ratio masks (IRMs) using simulated training pairs of clean and noisy speech data. Next, a student model is trained to learn an improved speech presence probability by incorporating the estimated IRMs from the teacher model into the IMCRA approach. The student model can be compactly designed in a causal processing mode having no latency with the guidance of a complex and noncausal teacher model. Moreover, the clean speech requirement, which is difficult to meet in real-world adverse environments, can be relaxed for training the student model, implying that noisy speech data can be directly used to adapt the regression-based enhancement model to further improve speech recognition accuracies for noisy speech collected in such conditions. Experiments on the CHiME-4 challenge task show that our best student model with bidirectional gated recurrent units (BGRUs) can achieve a relative word error rate (WER) reduction of 18.85% for the real test set when compared to unprocessed system without acoustic model retraining. However, the traditional teacher model degrades the performance of the unprocessed system in this case. In addition, the student model with a deep neural network (DNN) in causal mode having no latency yields a relative WER reduction of 7.94% over the unprocessed system with 670 times less computing cycles when compared to the BGRU-equipped student model. Finally, the conventional speech enhancement and IRM-based deep learning method destroyed the ASR performance when the recognition system became more powerful. While our proposed approach could still improve the ASR performance even in the more powerful recognition system.

Mixed-Bandwidth Cross-Channel Speech Recognition via Joint Optimization of DNN-Based Bandwidth Expansion and Acoustic Modeling

Automatic speech recognition (ASR) systems are often built using scene related speech data due to large variations of transmission channels and sampling rates in different scenarios. In this study, we propose a general framework that establishes a unified model for diversified speech data with different sampling rates and channels. The framework is a joint optimization of deep neural network (DNN)-based bandwidth expansion and acoustic modeling to exploit a large amount of diversified training data. First, we design two novel DNN architectures to map the acoustic features from narrowband to wideband speech through direct mapping and progressive mapping. The learning targets of the direct mapping DNN (DNN-DM) are the acoustic features extracted from speech with the largest bandwidth, while the acoustic features from speech with all the other bandwidths are used as input. A progressive stacking network (PSN) gradually maps the features from the low sampling rates to the highest sampling rate through the design of intermediate target layers via multitask training. Then, in addition to these bandwidth expansion networks, we investigate several joint training strategies for DNN-based acoustic models. Our experiments conducted on three diversified large-scale Mandarin speech datasets with different recording channels and sampling rates (6, 8, and 16 kHz) show that the proposed unified model using PSN for bandwidth expansion not only is a more flexible and compact design than conventional multiple acoustic models with each bandwidth for a specific sampling rate, but also yields consistent and significant improvements over bandwidth-dependent models with an average relative word error rate reduction of 6.2%, indicating that the proposed model can fully utilize the diversified cross-channel speech data with multiple bandwidths. Moreover, the proposed methods are verified to be robust on different realistic scenes and can be effectively extended to a long short-term memory framework.