检索结果-内蒙古大学图书馆

44th International ACM SIGIR Conference on Research and Development in Information Retrieval

作者： Pradeep, Ronak Ma, Xueguang Nogueira, Rodrigo Lin, Jimmy Univ Waterloo David R Cheriton Sch Comp Sci Waterloo ON Canada

ISBN: (纸本)9781450380379

The COVID-19 pandemic has brought about a proliferation of harmful news articles online, with sources lacking credibility and misrepresenting scientific facts. Misinformation has real consequences for consumer health search, i.e., users searching for health information. In the context of multi-stage ranking architectures, there has been little work exploring whether they prioritize correct and credible information over misinformation. We find that, indeed, training models on standard relevance ranking datasets like MS MARCO passage-which have been curated to contain mostly credible information-yields models that might also promote harmful misinformation. To rectify this, we propose a label prediction technique that can separate helpful from harmful content. Our design leverages pretrained sequence-to-sequence transformer models for both relevance ranking and label prediction. Evaluated at the TREC 2020 Health Misinformation Track, our techniques represent the top-ranked system: Our best submitted run was 19.2 points higher than the second-best run based on the primary metric, a 68% relative improvement. Additional post-hoc experiments show that we can boost effectiveness by another 3.5 points.

关键词： Multi-Stage Ranking sequence-to-sequence models

来源：评论

学校读者我要写书评

暂无评论

JOINT MULTI-PITCH DETECTION AND SCORE TRANSCRIPTION FOR POLYPHONIC PIANO MUSIC

JOINT MULTI-PITCH DETECTION AND SCORE TRANSCRIPTION FOR POLY...

引用

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

作者： Liu, Lele Morfi, Veronica Benetos, Emmanouil Queen Mary Univ London Ctr Digital Mus London England

ISBN: (纸本)9781728176055

Research on automatic music transcription has largely focused on multi-pitch detection;there is limited discussion on how to obtain a machine- or human-readable score transcription. In this paper, we propose a method for joint multi-pitch detection and score transcription for polyphonic piano music. The outputs of our system include both a piano-roll representation (a descriptive transcription) and a symbolic musical notation (a prescriptive transcription). Unlike traditional methods that further convert MIDI transcriptions into musical scores, we use a multitask model combined with a Convolutional Recurrent Neural Network and sequence-to-sequence models with attention mechanisms. We propose a Reshaped score representation that outperforms a LilyPond representation in terms of both prediction accuracy and time/memory resources, and compare different input audio spectrograms. We also create a new synthesized dataset for score transcription research. Experimental results show that the joint model outperforms a single-task model in score transcription.

关键词： Automatic music transcription sequence-to-sequence models score transcription

来源：评论

学校读者我要写书评

暂无评论

Improving RNN-T ASR Accuracy Using Context Audio 22

Improving RNN-T ASR Accuracy Using Context Audio

引用

Interspeech Conference

作者： Schwarz, Andreas Sklyar, Ilya Wiesler, Simon Amazon Alexa Aachen Germany

ISBN: (纸本)9781713836902

We present a training scheme for streaming automatic speech recognition (ASR) based on recurrent neural network transducers (RNN-T) which allows the encoder network to learn to exploit context audio from a stream, using segmented or partially labeled sequences of the stream during training. We show that the use of context audio during training and inference can lead to word error rate reductions of more than 6% in a realistic production setting for a voice assistant ASR system. We investigate the effect of the proposed training approach on acoustically challenging data containing background speech and present data points which indicate that this approach helps the network learn both speaker and environment adaptation. To gain further insight into the ability of a long short-term memory (LSTM) based ASR encoder to exploit long-term context, we also visualize RNN-T loss gradients with respect to the input.

关键词： ASR sequence-to-sequence models RNN-T acoustic models

来源：评论

学校读者我要写书评

暂无评论

Context-Dependent sequence-to-sequence Turkish Spelling Correction

引用

ACM TRANSACTIONS ON ASIAN AND LOW-RESOURCE LANGUAGE INFORMATION PROCESSING 2020年第4期19卷 1–16页

作者： Buyuk, Osman Kocaeli Univ Umuttepe Yerleskesi TR-41380 Kocaeli Turkey Sestek Speech Enabled Software Technol Inc TR-34396 Istanbul Turkey

In this article, we make use of sequence-to-sequence (seq2seq) models for spelling correction in the agglutinative Turkish language. In the baseline system, misspelled and target words are split into their letters and the letter sequences are fed into the seq2seq model. We prefer letters as the unit of the model due to the agglutinative nature of Turkish, which results in an impractical dictionary size when words are used as a dictionary unit. In order to improve the baseline performance, we incorporate right and left context of the misspelled words. All context words are represented with their first three consonants in the context-dependent model. We train the seq2seq models using a large text corpus collected automatically from the Internet. The corpus contains approximately 4 million sentences. We randomly introduce substitution, deletion, and insertion spelling errors to the words in the corpus. We test the performance of the proposed context-dependent seq2seq model using synthetic and realistic test sets. The synthetic test set is constructed similar to the training set. The realistic test set contains human-made misspellings from Twitter messages. In the experiments, we observed that the proposed context-dependent model performs significantly better than the baseline system. Its correction accuracy reaches 94% on the synthetic dataset. Additionally, the proposed method provides 2.1% absolute improvement over a state-of-the-art Turkish spelling correction system on the Twitter test set.

关键词： Spelling correction sequence-to-sequence models agglutinative language

来源：评论

学校读者我要写书评

暂无评论

Investigating alignment interpretability for low-resource NMT

引用

MACHINE TRANSLATION 2020年第4期34卷 305-323页

作者： Boito, Marcely Zanon Villavicencio, Aline Besacier, Laurent Univ Grenoble Alpes Lab Informat Grenoble Grenoble France Univ Sheffield Dept Comp Sci Sheffield S Yorkshire England Univ Fed Rio Grande do Sul Inst Informat Porto Alegre RS Brazil

The attention mechanism in Neural Machine Translation (NMT) models added flexibility to translation systems, and the possibility to visualize soft-alignments between source and target representations. While there is much debate about the relationship between attention and the yielded output for neural models (Jain and Wallace 2019;Serrano and Smith 2019;Wiegreffe and Pinter 2019;Vashishth et al. 2019), in this paper we propose a different assessment, investigating soft-alignment interpretability in low-resource scenarios. We experimented with different architectures (RNN (Bahdanau et al. 2015), 2D-CNN (Elbayad et al. 2018), and Transformer (Vaswani et al. 2017)), comparing them with regards to their ability to produce directly exploitable alignments. For evaluating exploitability, we replicated the Unsupervised Word Segmentation (UWS) task from Godard et al. (2018). There, source words are translated into unsegmented phone sequences. Posterior to training, the resulting soft-alignments are used for producing segmentation over the target side. Our results showed that a RNN-based NMT model produced the most exploitable alignments in this scenario. We then investigated methods for increasing its UWS scores by comparing the following methodologies: monolingual pre-training, input representation augmentation (hybrid model), and explicit word length optimization during training. We reached the best results by using the hybrid model, which uses an intermediate monolingual-rooted segmentation from a non-parametric Bayesian model (Goldwater 2007) to enrich the input representation before training.

关键词： Low-resource languages Attention mechanism sequence-to-sequence models Unsupervised word segmentation Computational language documentation Neural machine translation

来源：评论

学校读者我要写书评

暂无评论

Candidate fusion: Integrating language modelling into a sequence-to-sequence handwritten word recognition architecture

引用

PATTERN RECOGNITION 2021年 112卷 107790-107790页

作者： Kang, Lei Riba, Pau Villegas, Mauricio Fornes, Alicia Rusinol, Marcal Univ Autonoma Barcelona Dept Comp Sci Comp Vis Ctr Barcelona Spain Omni Us Berlin Germany

sequence-to-sequence models have recently become very popular for tackling handwritten word recognition problems. However, how to effectively integrate an external language model into such recognizer is still a challenging problem. The main challenge while training a language model is to deal with the language model corpus which is usually different to the one used for training the handwritten word recognition system. Thus, the bias between both word corpora leads to incorrectness on the transcriptions, providing similar or even worse performances on the recognition task. In this work, we introduce Candidate Fusion, a novel way to integrate an external language model to a sequence-to-sequence architecture. Moreover, it provides suggestions from an external language knowledge, as a new input to the sequence-to-sequence recognizer. Hence, Candidate Fusion provides two improvements. On the one hand, the sequence-to-sequence recognizer has the flexibility to not only combine the information from itself and the language model, but also choose the importance of the information provided by the language model. On the other hand, the external language model has the ability to adapt itself to the training corpus and even learn the most common errors produced from the recognizer. Finally, by conducting comprehensive experiments, the Candidate Fusion proves to outperform the state-of-the-art language models for handwritten word recognition tasks. (c) 2020 Elsevier Ltd. All rights reserved.

关键词： Handwritten word recognition sequence-to-sequence models Language model Candidate fusion

来源：评论

学校读者我要写书评

暂无评论

Weakly supervised action segmentation with effective use of attention and self-attention

引用

COMPUTER VISION AND IMAGE UNDERSTANDING 2021年 213卷 103298-103298页

作者： Bin Ng, Yan Fernando, Basura ASTAR Inst High Performance Comp 1 Fusionopolis Way 08 Connexis North Singapore 138632 Singapore

This paper generates human action sequences using a novel hybrid sequence-to-sequence model that outputs a sequence of actions in the chronological order of the actions being performed in the longer activity of a given video. At test time, our models are able to generate action for each frame using weak supervision. We evaluate several sequence-to-sequence models to solve this task and demonstrate that they are able to solve action segment generation tasks on three challenging action recognition datasets. We present how to use self attention and standard attention mechanisms with known sequence-to-sequence models for weakly supervised video action segmentation. Our new architecture is effective for weakly supervised action segmentation that uses a combination of recurrent and transformer-based sequence-to-sequence models. Our architecture consists of Transformers and GRU encoders to encode temporal information and we use self-attention and standard attention during the decoding process. We introduce an effective positional weight prior to further improve action segmentation performance. Using this architecture and two types of attention along with positional weight priors, we obtain state-of-the-art results on Breakfast and 50Salads datasets for weakly supervised action segmentation.

关键词： Weakly supervised action segmentation Self-attention sequence-to-sequence models

来源：评论

学校读者我要写书评

暂无评论

Low-Latency sequence-to-sequence Speech Recognition and Translation by Partial Hypothesis Selection 21

Low-Latency Sequence-to-Sequence Speech Recognition and Tran...

引用

Interspeech Conference

作者： Liu, Danni Spanakis, Gerasimos Niehues, Jan Maastricht Univ Dept Data Sci & Knowledge Engn Maastricht Netherlands

ISBN: (纸本)9781713820697

Encoder-decoder models provide a generic architecture for sequence-to-sequence tasks such as speech recognition and translation. While offline systems are often evaluated on quality metrics like word error rates (WER) and BLEU scores, latency is also a crucial factor in many practical use-cases. We propose three latency reduction techniques for chunk-based incremental inference and evaluate their accuracy-latency trade-off. On the 300-hour How2 dataset, we reduce latency by 83% to 0.8 second by sacrificing 1% WER (6% rel.) compared to offline transcription. Although our experiments use the Transformer, the partial hypothesis selection strategies are applicable to other encoder-decoder models. To reduce expensive re-computation as new chunks arrive, we propose to use a unidirectionally-attending encoder. After an adaptation procedure to partial sequences, the unidirectional model performs on-par with the original model. We further show that our approach is also applicable to speech translation. On the How2 English-Portuguese speech translation dataset, we reduce latency to 0.7 second (-84% rel.) while incurring a loss of 2.4 BLEU points (5% rel.) compared to the offline system.

关键词： low-latency sequence-to-sequence models speech recognition speech translation

来源：评论

学校读者我要写书评

暂无评论

Distilling the Knowledge of BERT for sequence-to-sequence ASR 21

Distilling the Knowledge of BERT for Sequence-to-Sequence AS...

引用

Interspeech Conference

作者： Futami, Hayato Inaguma, Hirofumi Ueno, Sei Mimura, Masato Sakai, Shinsuke Kawahara, Tatsuya Kyoto Univ Grad Sch Informat Sakyo Ku Kyoto Japan

ISBN: (纸本)9781713820697

Attention-based sequence-to-sequence (seq2seq) models have achieved promising results in automatic speech recognition (ASR). However, as these models decode in a left-to-right way, they do not have access to context on the right. We leverage both left and right context by applying BERT as an external language model to seq2seq ASR through knowledge distillation. In our proposed method, BERT generates soft labels to guide the training of seq2seq ASR. Furthermore, we leverage context beyond the current utterance as input to BERT. Experimental evaluations show that our method significantly improves the ASR performance from the seq2seq baseline on the Corpus of Spontaneous Japanese (CSJ). Knowledge distillation from BERT outperforms that from a transformer LM that only looks at left context. We also show the effectiveness of leveraging context beyond the current utterance. Our method outperforms other LM application approaches such as n-best rescoring and shallow fusion, while it does not require extra inference cost.

关键词： speech recognition sequence-to-sequence models language model BERT knowledge distillation

来源：评论

学校读者我要写书评

暂无评论

Speech To Semantics: Improve ASR and NLU Jointly via All-Neural Interfaces 21

Speech To Semantics: Improve ASR and NLU Jointly via All-Neu...

引用

Interspeech Conference

作者： Rao, Milind Raju, Anirudh Dheram, Pranav Bui, Bach Rastrow, Ariya Amazon Alexa Seattle WA 98109 USA

ISBN: (纸本)9781713820697

We consider the problem of spoken language understanding (SLU) of extracting natural language intents and associated slot arguments or named entities from speech that is primarily directed at voice assistants. Such a system subsumes both automatic speech recognition (ASR) as well as natural language understanding (NLU). An end-to-end joint SLU model can be built to a required specification opening up the opportunity to deploy on hardware constrained scenarios like devices enabling voice assistants to work offline, in a privacy preserving manner, whilst also reducing server costs. We first present models that extract utterance intent directly from speech without intermediate text output. We then present a compositional model, which generates the transcript using the Listen Attend Spell ASR system and then extracts interpretation using a neural NLU model. Finally, we contrast these methods to a jointly trained end-to-end joint SLU model, consisting of ASR and NLU subsystems which are connected by a neural network based interface instead of text, that produces transcripts as well as NLU interpretation. We show that the jointly trained model shows improvements to ASR incorporating semantic information from NLU and also improves NLU by exposing it to ASR confusion encoded in the hidden layer.

关键词： Speech Recognition Spoken Language Understanding sequence-to-sequence models Multitask 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：