检索结果-内蒙古大学图书馆

Interspeech Conference

作者： Zhou, Kun Sisman, Berrak Li, Haizhou Natl Univ Singapore NUS Dept ECE Singapore Singapore Singapore Univ Technol & Design SUTD Singapore Singapore

ISBN: (纸本)9781713836902

Emotional voice conversion (EVC) aims to change the emotional state of an utterance while preserving the linguistic content and speaker identity. In this paper, we propose a novel 2-stage training strategy for sequence-to-sequence emotional voice conversion with a limited amount of emotional speech data. We note that the proposed EVC framework leverages text-to-speech (TTS) as they share a common goal that is to generate high-quality expressive voice. In stage 1, we perform style initialization with a multi-speaker TTS corpus, to disentangle speaking style and linguistic content. In stage 2, we perform emotion training with a limited amount of emotional speech data, to learn how to disentangle emotional style and linguistic information from the speech. The proposed framework can perform both spectrum and prosody conversion and achieves significant improvement over the state-of-the-art baselines in both objective and subjective evaluation.

关键词： Emotional voice conversion sequence-to-sequence limited data

来源：评论

学校读者我要写书评

暂无评论

sequence-to-sequence SINGING SYNTHESIS USING THE FEED-FORWARD TRANSFORMER

SEQUENCE-TO-SEQUENCE SINGING SYNTHESIS USING THE FEED-FORWAR...

引用

IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Blaauw, Merlijn Bonada, Jordi Univ Pompeu Fabra Mus Technol Grp Barcelona Spain

ISBN: (纸本)9781509066315

We propose a sequence-to-sequence singing synthesizer, which avoids the need for training data with pre-aligned phonetic and acoustic features. Rather than the more common approach of a content-based attention mechanism combined with an autoregressive decoder, we use a different mechanism suitable for feed-forward synthesis. Given that phonetic timings in singing are highly constrained by the musical score, we derive an approximate initial alignment with the help of a simple duration model. Then, using a decoder based on a feed-forward variant of the Transformer model, a series of self-attention and convolutional layers refines the result of the initial alignment to reach the target acoustic features. Advantages of this approach include faster inference and avoiding the exposure bias issues that affect autoregressive models trained by teacher forcing. We evaluate the effectiveness of this model compared to an autoregressive baseline, the importance of self-attention, and the importance of the accuracy of the duration model.

关键词： Singing synthesis sequence-to-sequence self-attention feed-forward transformer

来源：评论

学校读者我要写书评

暂无评论

Basecalling Using Joint Raw and Event Nanopore Data sequence-to-sequence Processing

引用

SENSORS 2022年第6期22卷 2275页

作者： Napieralski, Adam Nowak, Robert Warsaw Univ Technol Fac Elect & Informat Technol Inst Comp Sci PL-00665 Warsaw Poland

Third-generation DNA sequencers provided by Oxford Nanopore Technologies (ONT) produce a series of samples of an electrical current in the nanopore. Such a time series is used to detect the sequence of nucleotides. The task of translation of current values into nucleotide symbols is called basecalling. Various solutions for basecalling have already been proposed. The earlier ones were based on Hidden Markov Models, but the best ones use neural networks or other machine learning models. Unfortunately, achieved accuracy scores are still lower than competitive sequencing techniques, like Illumina's. Basecallers differ in the input data type-currently, most of them work on a raw data straight from the sequencer (time series of current). Still, the approach of using event data is also explored. Event data is obtained by preprocessing of raw data and dividing it into segments described by several features computed from raw data values within each segment. We propose a novel basecaller that uses joint processing of raw and event data. We define basecalling as a sequence-to-sequence translation, and we use a machine learning model based on an encoder-decoder architecture of recurrent neural networks. Our model incorporates twin encoders and an attention mechanism. We tested our solution on simulated and real datasets. We compare the full model accuracy results with its components: processing only raw or event data. We compare our solution with the existing ONT basecaller-Guppy. Results of numerical experiments show that joint raw and event data processing provides better basecalling accuracy than processing each data type separately. We implement an application called Ravvent, freely available under MIT licence.

关键词： basecalling sequence-to-sequence nanopore bioinformatics encoder decoder attention neural network machine learning joint processing

来源：评论

学校读者我要写书评

暂无评论

PREDICTION OF SIGNIFICANT WAVE HEIGHT BASED ON GATED RECURRENT UNIT AND sequence-to-sequence NETWORKS IN THE TAIWAN STRAIT

引用

COMPUTING AND INFORMATICS 2022年第3期41卷 885-904页

作者： Zeng, Yindong Ma, Jinshan Chen, Xinping Guo, Minquan Ren, Zaichang Jiang, Yanqi Zhang, Zhenchang Ocean Univ China Coll Ocean & Atmospher Sci Qingdao 266071 Peoples R China Marine Forecasting Ctr Fujian Prov Fuzhou 350003 Peoples R China Fujian Agr & Forestry Univ Inst Comp Applicat Fuzhou 350000 Peoples R China Minist Nat Resources China Natl Marine Hazard Mitigat Serv Beijing 100194 Peoples R China Hohai Univ Coll Harbour Coastal & Offshore Engn Nanjing 210098 Peoples R China

Wave forecasting approaches based on deep learning techniques have recently made a great progress. In this study, we developed a deep learning model based on Gated Recurrent Unit (GRU) and sequence-to-sequence neural networks (GRUS), to improve the forecasting accuracy of significant wave heights for the Taiwan Strait, where ocean waves and winds own their unique characteristics. The performances of our proposed GRUS model and the other deep learning models based onWaveNet and Long Short-Term Memory (LSTM) were compared by means of wind and wave observations at three buoys in the study area. Model parameters were optimized by means of various model experiments. Performance comparison illustrates that our proposed GRUS model outperforms the other models in 24-hour Hs forecasting, while the GRUS has extraordinary ability for short-term prediction (prediction horizon is less than 6 h). Moreover, for high wave states prediction (e.g., wave height over 4 m), the GRUS has the strongest prediction ability among the models, in which forecasted wave heights are mostly lower than the corresponding observations.

关键词： Wave forecasting significant wave height gated recurrent unit long short-term memory sequence-to-sequence

来源：评论

学校读者我要写书评

暂无评论

A sequence-to-sequence remaining useful life prediction method combining unsupervised LSTM encoding-decoding and temporal convolutional network

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2022年第8期33卷 085013-085013页

作者： Li, Jialin Chen, Renxiang Huang, Xianzhen Chongqing Jiaotong Univ Chongqing Engn Lab Transportat Engn Applicat Robo Chongqing Peoples R China Northeastern Univ Sch Mech Engn & Automat Shenyang Peoples R China

Remaining useful life (RUL) prediction methods based on deep neural networks (DNNs) have received much attention in recent years. The collected time-series signals are usually processed by the sliding time window method into several segments with the same sequence length as the input. However, the signal processing is not only time-consuming, but also relies too much on personal experience. Moreover, the length of the time window affects the prediction results and the prediction range. Obviously, it is more desirable to remove the data processing and use an entire time series signal as the input for predicting the RUL, i.e. sequence-to-sequence RUL prediction. In order to remove the shortcomings of signal processing, this paper uses a long short-term memory (LSTM) and encoding-decoding framework to construct an unsupervised sequence data processing model. Then, a temporal convolutional network, based on a convolutional neural network, is used to further process the output data of the unsupervised sequence data processing model. The proposed sequence-to-sequence RUL prediction method not only maintains the complete sequence of the data, but has a good capability for data processing. The open access C-MAPSS simulation datasets are used for validation. The validation results show that the proposed method can realize unsupervised sequence signal reconstruction. Moreover, it has better prediction results and prediction efficiency.

关键词： remaining useful life prediction sequence-to-sequence temporal convolutional network long short-term memory network

来源：评论

学校读者我要写书评

暂无评论

Emotionally-Aware sequence-to-sequence Models for Conversational Systems 16th

Emotionally-Aware Sequence-to-Sequence Models for Conversati...

引用

16th International Conference on Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP 2020 in conjunction with the 13th International Conference on Frontiers of Information Technology, Applications and Tools, FITAT 2020

作者： Tran-Luong, Quoc-Dai Le, Anh-Cuong Pham, Duc-Hong Natural Language Processing and Knowledge Discovery Laboratory Faculty of Information Technology Ton Duc Thang University Ho Chi Minh City Viet Nam Faculty of Information Technology Electric Power University Hanoi Viet Nam

ISBN: (纸本)9789813367562

Emotion is an important factor to the success of conversational agents or chatbot. Existing chatbots have been shown to be good for generating natural responses. However, they still lack mechanisms to integrate emotional information. In this paper, we propose an approach that can produce emotion-specific responses by encoding emotional information into sequence-to-sequence model. The experimental results show that the proposed model generates responses appropriate to content and emotion. Specifically, this model improves up to 10% in emotional precision and 50% BLEU score compared to the baseline. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021.

关键词： Bert Chatbot Conversational agent Emotion sequence-to-sequence

来源：评论

学校读者我要写书评

暂无评论

Learning sequence-to-sequence affinity metric for near-online multi-object tracking

引用

KNOWLEDGE AND INFORMATION SYSTEMS 2020年第10期62卷 3911-3930页

作者： Feng, Weijiang Lan, Long Zhang, Xiang Luo, Zhigang Natl Univ Def Technol Coll Comp Changsha Peoples R China Natl Univ Def Technol Coll Comp Inst Quantum Informat Changsha Peoples R China Natl Univ Def Technol Coll Comp State Key Lab High Performance Comp Changsha Peoples R China

In this paper, we propose a sequence-to-sequence affinity metric for the data association of near-online multi-object tracking. The proposed metric learns the affinity between track sequence consisting of the already associated detections and hypothesis sequence consisting of detections in the near future. With the potential hypothesis sequences, we leverage the idea that if a track sequence has a high affinity for a hypothesis sequence, and the hypothesis sequence also shares a close affinity for a current detection, then the affinity between the track sequence and the detection is high. By using the short hypothesis sequence as a "bridge", the proposed sequence-to-sequence affinity metric enhances the conventional track sequence to detection affinity metric and improves its robustness to object occlusion and missing. Besides, in order to eliminate the negative effects of false alarms, we propose a false alarm model using both appearance and scale features of detection. The robustness of the proposed affinity metric allows us to use a simple greedy data association algorithm. Experimental results on the challenging MOT16 and MOT17 benchmarks demonstrate the effectiveness of our method.

关键词： Multi-object tracking sequence-to-sequence MOT Challenge

来源：评论

学校读者我要写书评

暂无评论

sequence-to-sequence Acoustic Modeling for Voice Conversion

引用

IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING 2019年第3期27卷 631-644页

作者： Zhang, Jing-Xuan Ling, Zhen-Hua Liu, Li-Juan Jiang, Yuan Dai, Li-Rong Univ Sci & Technol China Natl Engn Lab Speech & Language Informat Proc Hefei 230027 Anhui Peoples R China iFLYTEK Co Ltd Hefei 230088 Anhui Peoples R China

In this paper, a neural network named sequence-to-sequence ConvErsion NeTwork (SCENT) is presented for acoustic modeling in voice conversion. At training stage, a SCENT model is estimated by aligning the feature sequences of source and target speakers implicitly using attention mechanism. At the conversion stage, acoustic features and durations of source utterances are converted simultaneously using the unified acoustic model. Mel-scale spectrograms are adopted as acoustic features, which contain both excitation and vocal tract descriptions of speech signals. The bottleneck features extracted from source speech using an automatic speech recognition model are appended as an auxiliary input. A WaveNet vocoder conditioned on Mel-spectrograms is built to reconstruct waveforms from the outputs of the SCENT model. It is worth noting that our proposed method can achieve appropriate duration conversion, which is difficult in conventional methods. Experimental results show that our proposed method obtained better objective and subjective performance than the baseline methods using Gaussian mixture models and deep neural networks as acoustic models. This proposed method also outperformed our previous work, which achieved the top rank in Voice Conversion Challenge 2018. Ablation tests further confirmed the effectiveness of several components in our proposed method.

关键词： Voice conversion sequence-to-sequence attention Mel-spectrogram

来源：评论

学校读者我要写书评

暂无评论

High Performance sequence-to-sequence Model for Streaming Speech Recognition 21

High Performance Sequence-to-Sequence Model for Streaming Sp...

引用

Interspeech Conference

作者： Thai-Son Nguyen Ngoc-Quan Pham Stueker, Sebastian Waibel, Alex Karlsruhe Inst Technol Inst Anthropomat & Robot Karlsruhe Germany

ISBN: (纸本)9781713820697

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 the bidirectional LSTM (BLSTM). In this paper, we propose several techniques to mitigate these problems. We introduce an additional loss function controlling the uncertainty of the attention mechanism, a modified beam search identifying partial, stable hypotheses, ways of working with BLSTM in the encoder, and the use of chunked BLSTM. Our experiments show that with the right combination of these techniques, it is possible to perform run-on speech recognition with low word-based latency without sacrificing in word error rate performance.

关键词： sequence-to-sequence online streaming

来源：评论

学校读者我要写书评

暂无评论

Enhancing sequence-to-sequence Text-to-Speech with Morphology 21

Enhancing Sequence-to-Sequence Text-to-Speech with Morpholog...

引用

Interspeech Conference

作者： Taylor, Jason Richmond, Korin Univ Edinburgh Ctr Speech Technol Res Edinburgh Midlothian Scotland

ISBN: (纸本)9781713820697

Neural sequence-to-sequence (S2S) modelling encodes a single, unified representation for each input sequence. When used for text-to-speech synthesis (TTS), such representations must embed ambiguities between English spelling and pronunciation. For example, in pothole and there the character sequence th sounds different. This can be problematic when predicting pronunciation directly from letters. We posit pronunciation becomes easier to predict when letters are grouped into subword units like morphemes (e.g. a boundary lies between t and h in pothole but not there). Moreover, morphological boundaries can reduce the total number of, and increase the counts of, seen unit subsequences. Accordingly, we test here the effect of augmenting input sequences of letters with morphological boundaries. We find morphological boundaries substantially lower the Word and Phone Error Rates (WER and PER) for a Bi-LSTM performing G2P on one hand, and also increase the naturalness scores of Tacotrons performing TTS in a MUSHRA listening test on the other. The improvements to TTS quality are such that grapheme input augmented with morphological boundaries outperforms phone input without boundaries. Since morphological segmentation may be predicted with high accuracy, we highlight this simple pre-processing step has important potential for S2S modelling in TTS.

关键词： Speech Synthesis sequence-to-sequence Morphology Pronunciation

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：