A Pruned Rnnlm Lattice-Rescoring Algorithm for Automatic Speech Recognition

Xu, Hainan; Chen, Tongfei; Gao, Dongji; Wang, Yiming; Li, Ké; Goel, Nagendra Kumar; Carmiel, Yishay; Povey, Daniel; Khudanpur, Sanjeev

doi:10.1109/icassp.2018.8461974

Cited by 95 publications

(76 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Proposed+pos performs the best on LectureSS, while proposed performs the best on SEAME 3 . It indicates that POS features help generator generate more useful code-switching sentences on LectureSS, but not SEAME.…”

Section: Language Modelingmentioning

confidence: 99%

Code-Switching Sentence Generation by Generative Adversarial Networks and its Application to Data Augmentation

Chang¹,

Chuang²,

Lee³

2019

Interspeech 2019

View full text Add to dashboard Cite

Code-switching is about dealing with alternative languages in speech or text. It is partially speaker-dependent and domainrelated, so completely explaining the phenomenon by linguistic rules is challenging. Compared to most monolingual tasks, insufficient data is an issue for code-switching. To mitigate the issue without expensive human annotation, we proposed an unsupervised method for code-switching data augmentation. By utilizing a generative adversarial network, we can generate intra-sentential code-switching sentences from monolingual sentences. We applied the proposed method on two corpora, and the result shows that the generated code-switching sentences improve the performance of code-switching language models.

show abstract

Section: Language Modelingmentioning

confidence: 99%

Code-Switching Sentence Generation by Generative Adversarial Networks and its Application to Data Augmentation

Chang¹,

Chuang²,

Lee³

2019

Interspeech 2019

View full text Add to dashboard Cite

show abstract

“…For NN-LM rescoring, we followed the setup described in [27], which directly rescored on the lattice; the word embedding dimension used in our setup is 1024. For the training data sets, in addition to Switchboard and Fisher transcripts, the out-domain Washington conversational Web corpus (191M words) was used.…”

Section: Language Modelmentioning

confidence: 99%

Automatic Speech Recognition System with Output-Gate Projected Gated Recurrent Unit

Cheng

Zhang

2019

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

The long short-term memory recurrent neural network (LSTM) has achieved tremendous success for automatic speech recognition (ASR). However, the complicated gating mechanism of LSTM introduces a massive computational cost and limits the application of LSTM in some scenarios. In this paper, we describe our work on accelerating the decoding speed and improving the decoding accuracy. First, we propose an architecture, which is called Projected Gated Recurrent Unit (PGRU), for ASR tasks, and show that the PGRU can consistently outperform the standard GRU. Second, to improve the PGRU generalization, particularly on large-scale ASR tasks, we propose the Output-gate PGRU (OPGRU). In addition, the time delay neural network (TDNN) and normalization methods are found beneficial for OPGRU. In this paper, we apply the OPGRU for both the acoustic model and recurrent neural network language model (RNN-LM). Finally, we evaluate the PGRU on the total Eval2000 / RT03 test sets, and the proposed OPGRU single ASR system achieves 0.9% / 0.9% absolute (8.2% / 8.6% relative) reduction in word error rate (WER) compared to our previous best LSTM single ASR system. Furthermore, the OPGRU ASR system achieves significant speed-up on both acoustic model and language model rescoring.

show abstract

“…In order to overcome this, two sets of sub-word units (7663 sub-word units generated by Morfessor 6 [26] and 9135 BPE sub-word units [25]) were used for training additional Transformer-XL LMs. RNN-LM was applied to rescore lattices [42], while the other NNLMs rescored 1000-best recognition hypotheses. Rescoring results for system #1 from Table 6 with different NNLMs are presented in Table 7.…”

Section: Final Acoustic Modelsmentioning

confidence: 99%

The STC ASR System for the VOiCES from a Distance Challenge 2019

Medennikov

Khokhlov²,

Romanenko

et al. 2019

Interspeech 2019

View full text Add to dashboard Cite

This paper is a description of the Speech Technology Center (STC) automatic speech recognition (ASR) system for the "VOiCES from a Distance Challenge 2019". We participated in the Fixed condition of the ASR task, which means that the only training data available was an 80-hour subset of the Lib-riSpeech corpus. The main difficulty of the challenge is a mismatch between clean training data and distant noisy development/evaluation data. In order to tackle this, we applied room acoustics simulation and weighted prediction error (WPE) dereverberation. We also utilized well-known speaker adaptation using x-vector speaker embeddings, as well as novel room acoustics adaptation with R-vector room impulse response (RIR) embeddings. The system used a lattice-level combination of 6 acoustic models based on different pronunciation dictionaries and input features. N-best hypotheses were rescored with 3 neural network language models (NNLMs) trained on both words and sub-word units. NNLMs were also explored for out-of-vocabulary (OOV) words handling by means of artificial texts generation. The final system achieved Word Error Rate (WER) of 14.7% on the evaluation data, which is the best result in the challenge.

show abstract

A Pruned Rnnlm Lattice-Rescoring Algorithm for Automatic Speech Recognition

Cited by 95 publications

References 7 publications

Code-Switching Sentence Generation by Generative Adversarial Networks and its Application to Data Augmentation

Code-Switching Sentence Generation by Generative Adversarial Networks and its Application to Data Augmentation

Automatic Speech Recognition System with Output-Gate Projected Gated Recurrent Unit

The STC ASR System for the VOiCES from a Distance Challenge 2019

Contact Info

Product

Resources

About