检索结果-内蒙古大学图书馆

International Symposium on Chinese Spoken language processing

作者： Bowen Pang Huan Zhao Gaosheng Zhang Xiaoyue Yang Yang Sun Li Zhang Qing Wang Lei Xie Audio Speech and Language Processing Group (ASLP@NPU) School of Computer Science Northwestern Polytechnical University (NPU) Xi’an China Shenzhen Transsion Holding Limited

ISBN: (纸本)9798350397970

This paper describes the TSUP team’s submission to the ISCSLP 2022 conversational short-phrase speaker diarization (CSSD) challenge which particularly focuses on short-phrase conversations with a new evaluation metric called conversational diarization error rate (CDER). In this challenge, we explore three kinds of typical speaker diarization systems, which are spectral clustering (SC) based diarization, target-speaker voice activity detection (TS-VAD) and end-to-end neural diarization (EEND) respectively. Our major findings are summarized as follows. First, the SC approach is more favored over the other two approaches under the new CDER metric. Second, tuning on hyperparameters is essential to CDER for all three types of speaker diarization systems. Specifically, CDER becomes smaller when the length of sub-segments setting longer. Finally, multi-system fusion through DOVER-LAP will worsen the CDER metric on the challenge data. Our submitted SC system eventually ranks the third place in the challenge.

关键词： Measurement Voice activity detection Error analysis Oral communication Tuning

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Question Answering via Answer Diversifying 29

Unsupervised Question Answering via Answer Diversifying

引用

29th International Conference on Computational Linguistics, COLING 2022

作者： Nie, Yuxiang Huang, Heyan Chi, Zewen Mao, Xian-Ling School of Computer Science and Technology Beijing Institute of Technology Beijing China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications Beijing China Beijing Institute of Technology Southeast Academy of Information Technology Fujian China

Unsupervised question answering is an attractive task due to its independence on labeled data. Previous works usually make use of heuristic rules as well as pre-trained models to construct data and train QA models. However, most of these works regard named entity (NE) as the only answer type, which ignores the high diversity of answers in the real world. To tackle this problem, we propose a novel unsupervised method by diversifying answers, named DiverseQA. Specifically, the proposed method is composed of three modules: data construction, data augmentation and denoising filter. Firstly, the data construction module extends the extracted named entity into a longer sentence constituent as the new answer span to construct a QA dataset with diverse answers. Secondly, the data augmentation module adopts an answer-type dependent data augmentation process via adversarial training in the embedding level. Thirdly, the denoising filter module is designed to alleviate the noise in the constructed data. Extensive experiments show that the proposed method outperforms previous unsupervised models on five benchmark datasets, including SQuADv1.1, NewsQA, TriviaQA, BioASQ, and DuoRC. Besides, the proposed method shows strong performance in the few-shot learning setting. © 2022 Proceedings - International Conference on Computational Linguistics, COLING. All rights reserved.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

CaTT-KWS: A Multi-stage Customized Keyword Spotting Framework based on Cascaded Transducer-Transformer

arXiv

引用

arXiv 2022年

作者： Yang, Zhanheng Sun, Sining Li, Jin Zhang, Xiaoming Wang, Xiong Ma, Long Xie, Lei Audio Speech and Language Processing Group School of Computer Science Northwestern Polytechnical University Xi’an China Tencent Technology Co. Ltd Beijing China

Customized keyword spotting (KWS) has great potential to be deployed on edge devices to achieve hands-free user experience. However, in real applications, false alarm (FA) would be a serious problem for spotting dozens or even hundreds of keywords, which drastically affects user experience. To solve this problem, in this paper, we leverage the recent advances in transducer and transformer based acoustic models and propose a new multi-stage customized KWS framework named Cascaded Transducer-Transformer KWS (CaTT-KWS), which includes a transducer based keyword detector, a frame-level phone predictor based force alignment module and a transformer based decoder. Specifically, the streaming transducer module is used to spot keyword candidates in audio stream. Then force alignment is implemented using the phone posteriors predicted by the phone predictor to finish the first stage keyword verification and refine the time boundaries of keyword. Finally, the transformer decoder further verifies the triggered keyword. Our proposed CaTT-KWS framework reduces FA rate effectively without obviously hurting keyword recognition accuracy. Specifically, we can get impressively 0.13 FA per hour on a challenging dataset, with over 90% relative reduction on FA comparing to the transducer based detection model, while keyword recognition accuracy only drops less than 2%. Copyright © 2022, The Authors. All rights reserved.

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

DESNet: A Multi-Channel Network for Simultaneous speech Dereverberation, Enhancement and Separation

DESNet: A Multi-Channel Network for Simultaneous Speech Dere...

引用

IEEE Spoken language Technology Workshop

作者： Yihui Fu Jian Wu Yanxin Hu Mengtao Xing Lei Xie Audio Speech and Language Processing Group (ASLP) School of Computer Science Northwestern Polytechnical University Xi’an China Speech and Language Group Microsoft

ISBN: (数字)9781728170664

ISBN: (纸本)9781728170671

In this paper, we propose a multi-channel network for simultaneous speech dereverberation, enhancement and separation (DESNet). To enable gradient propagation and joint optimization, we adopt the attentional selection mechanism of the multi-channel features, which is originally proposed in end-to-end unmixing, fixed-beamforming and extraction (E2E-UFE) structure. Furthermore, the novel deep complex convolutional recurrent network (DCCRN) is used as the structure of the speech unmixing and the neural network based weighted prediction error (WPE) is cascaded before-hand for speech dereverberation. We also introduce the staged SNR strategy and symphonic loss for the training of the network to further improve the final performance. Experiments show that in non-dereverberated case, the proposed DESNet outperforms DCCRN and most state-of-the-art structures in speech enhancement and separation, while in dereverberated scenario, DESNet also shows improvements over the cascaded WPE-DCCRN networks.

关键词： Training Neural networks speech recognition speech enhancement Feature extraction Task analysis Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

Wekws: A Production First Small-Footprint End-to-End Keyword Spotting Toolkit

Wekws: A Production First Small-Footprint End-to-End Keyword...

引用

International Conference on Acoustics, speech, and Signal processing (ICASSP)

作者： Jie Wang Menglong Xu Jingyong Hou Binbin Zhang Xiao-Lei Zhang Lei Xie Fuping Pan School of Marine Science and Technology Northwestern Polytechnical University Xi’an China WeNet Open Source Community Horizon Robotics Beijing China Audio Speech and Language Processing Group (ASLP@NPU) School of Computer Science Northwestern Polytechnical University Xi’an China

Keyword spotting (KWS) enables speech-based user interaction and gradually becomes an indispensable component of smart devices. Recently, end-to-end (E2E) methods have be-come the most popular approach for on-device KWS tasks. However, there is still a gap between the research and deployment of E2E KWS methods. In this paper, we introduce WeKws, a production-quality, easy-to-build, and convenient-to-be-applied E2E KWS toolkit. WeKws contains the implementations of several state-of-the-art backbone networks, making it achieve highly competitive results on three publicly available datasets. To make WeKws a pure E2E toolkit, we utilize a refined max-pooling loss to make the model learn the ending position of the keyword by itself, which significantly simplifies the training pipeline and makes WeKws very efficient to be applied in real-world scenarios. The toolkit is publicly available at https://***/wenet-e2e/wekws.

关键词： Training Pipelines Production Signal processing Benchmark testing Acoustics Task analysis

来源：评论

学校读者我要写书评

暂无评论

The NPU-ASLP System for The ISCSLP 2022 Magichub Code-Swiching ASR Challenge

The NPU-ASLP System for The ISCSLP 2022 Magichub Code-Swichi...

引用

International Symposium on Chinese Spoken language processing

作者： Yuhao Liang Peikun Chen Fan Yu Xinfa Zhu Tianyi Xu Yingying Gao Lei Xie Audio Speech and Language Processing Group (ASLP@NPU) School of Computer Science Northwestern Polytechnical University Xi’an China China Mobile Research Institute Beijing China

ISBN: (纸本)9798350397970

This paper describes our NPU-ASLP system submitted to the ISCSLP 2022 Magichub Code-Switching ASR Challenge. In this challenge, we first explore several popular end-to-end multilingual ASR architectures and training strategies, including bi-encoder, language-aware encoder (LAE) and mixture of experts (MoE). To improve our system’s language modeling ability, we further attempt the internal language model as well as the long context language model. Given the limited training data in the challenge, we further investigate data augmentation strategies, including speed perturbation, pitch shifting, speech codec, SpecAugment and synthetic data from text-to-speech (TTS). Finally, we explore ROVER-based score fusion to make full use of complementary hypotheses from different models. Our submitted system achieves 16.87% on mix error rate (MER) on the test set and comes to the 2nd place in the challenge ranking.

关键词： Training speech codecs Perturbation methods Training data speech recognition Transformers speech processing

来源：评论

学校读者我要写书评

暂无评论

Meta Representation Learning Method for Robust Speaker Verification in Unseen Domains

Meta Representation Learning Method for Robust Speaker Verif...

引用

International Conference on Acoustics, speech, and Signal processing (ICASSP)

作者： Jian-Tao Zhang Yan Song Jin Li Wu Guo Hao-Yu Song Ian McLoughlin National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei China iFLYTEK Research iFLYTEK Co. Ltd. Hefei China The Australian National University Australian ICT Cluster Singapore Institute of Technology Singapore

This paper presents a meta representation learning method for robust speaker verification (SV) in unseen domains. It is known that the existing embedding learning based SV systems may suffer from domain mismatch issues. To address this, we propose an episodic training procedure to compensate domain mismatch conditions at runtime. Specifically, episodes are constructed with domain balanced episodic sampling from two different domains, and a new domain alignment (DA) module is added besides the feature extractor (FE) and classifier to existing network structures. In each episodic training iteration, FE and DA modules are optimized separately with different objectives to improve the robustness of learning. Besides, a cross-domain inter-class alignment (CDICA) loss is proposed for improving the domain generalization ability. Experimental results on CNCeleb and VoxCeleb benchmarks demonstrate significant performance gains for unseen domains in SV.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The Span-Based Transformer Classifier for Named Entity Recognition

The Span-Based Transformer Classifier for Named Entity Recog...

引用

International Conference on Autonomous Unmanned Systems, ICAUS 2021

作者： Li, Yangfan Sun, Xin Li, Hongao Yang, Kaige Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications School of Computer Science and Technology Beijing Institute of Technology Beijing100081 China Beijing Institute of Technology Southeast Academy of Information Technology Fujian Putian351100 China

ISBN: (纸本)9789811694912

Named entity recognition is a basic task in the field of natural language processing. The existing named entity recognition model has the problem of poor recognition effect on specific domain data sets. Aiming at the inconsistency of sequence labeling caused by the shortcomings of the traditional BIO (B-begin, I-inside, O-outside) annotation method, a named entity recognition model based on transformer is proposed, which uses the feature extraction of encoder in transformer and the labeling method of limited interval. Firstly, the BERT pre-training model is used to perform word-level dynamic vector representation of the input text. Then the vector representation of the text is intercepted by the span-based labeling method, and the feature of text and candidate named entity is extracted by self-attention mechanism. The results of experiment on the dataset in the industrial informatization field show the effectiveness of the model. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Dynamic Regret with Unknown Delays

SSRN

引用

SSRN 2023年

作者： Wu, Ping Liu, Zhengyang Lu, Haolin Huang, Heyan School of Computer Science and Technology Beijing Institute of Technology Beijing100081 China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications China

We study the problem of online convex optimization in dynamic environments with unknown and arbitrary delays. Without delays, existing work has achieved the optimal $\Theta(\sqrt{(1+\Delta_T)T})$ regret bound for any given sequence of comparators, where $T$ is the time horizon and $\Delta_T$ is the path-length of the comparator sequence. However, their algorithm assumes that the loss function can be revealed immediately to the agent after making the decision. In this paper, we propose a novel algorithm to tackle the delayed setting. We show that it enjoys $\mathcal{O}(\max\{\sqrt{(1+\Delta_T)D_T},(\hat{d}-1)\Delta_T\})$ regret bound, where $\hat{d}$ is the maximum delay and $D_T$ is the cumulative delay over the whole horizon. To complement our algorithm, we also establish an $\Omega(\sqrt{(1+\Delta_T)D_T})$ lower bound. The basic idea is to employ a delayed meta algorithm to track parallel delayed expert algorithms with varying step sizes. We further extend our algorithm to the case that we have no idea about the time horizon and cumulative delay using a novel doubling trick. Finally, numerical simulations are provided to illustrate the validity of our theoretical results. © 2023, The Authors. All rights reserved.

关键词： Comparators (optical)

来源：评论

学校读者我要写书评

暂无评论

DESIGN AND DEVELOPMENT OF MORPHOLOGICAL ANALYZER FOR TIGRIGNA VERBS USING HYBRID APPROACH

Research Square

引用

Research Square 2023年

作者： Gebremeskel, Hagos Gebremedhin Chong, Feng Heyan, Huang Beijing Engineering Research Center of High Volume Language Information Processing Cloud Computing Applications Laboratory School of Computer Science and Technology Beijing Institute of Technology China Department of Computer Science Mekelle University Tigray Mekelle231 Ethiopia

Morphological analyzer is the base for various high-level NLP applications such as information retrieval, spell checking, grammar checking, machine translation, speech recognition, POS tagging and automatic sentence construction. This paper is carefully designed for design and analysis of morphological analyzer Tigrigna verbs using hybrid of memory learning and rules based approaches. The experiment have conducted using Python 3 where TiMBL algorithms IB2 and TRIBL2, and Finite State Transducer rules are used. The performance of the system has been evaluated using 10 fold cross validation technique. Testing was conducted using optimized parameter settings for regular verbs and linguistic rules of the Tigrigna language allomorph and phonology for the irregular verbs. The accuracy of the memory based approach with optimized parameters of TiMBL algorithm IB2 and TRIBL2 was 93.24% and 92.31%, respectively. Finally, the hybrid approach had an actual performance of 95.6% using linguistic rules for handling irregular and copula verbs. © 2023, CC BY.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：