检索结果-内蒙古大学图书馆

learning a bidirectional mapping between human whole-body motion and natural language using deep recurrent neural networks

引用

ROBOTICS AND AUTONOMOUS SYSTEMS 2018年 109卷 13-26页

作者： Plappert, Matthias Mandery, Christian Asfour, Tamim Karlsruhe Inst Technol High Performance Humanoid Technol H2T Adenauerring 2Bldg 50-20 D-76131 Karlsruhe Germany

Linking human whole-body motion and natural language is of great interest for the generation of semantic representations of observed human behaviors as well as for the generation of robot behaviors based on natural language input. While there has been a large body of research in this area, most approaches that exist today require a symbolic representation of motions (e.g. in the form of motion primitives), which have to be defined a-priori or require complex segmentation algorithms. In contrast, recent advances in the field of neural networks and especially deep learning have demonstrated that sub-symbolic representations that can be learfied end-to-end usually outperform more traditional approaches, for applications such as machine translation. In this paper we propose a generative model that learns a bidirectional mapping between human whole-body motion and natural language using deep recurrent neural networks (RNNs) and sequence-to-sequence learning. Our approach does not require any segmentation or manual feature engineering and learns a distributed representation, which is shared for all motions and descriptions. We evaluate our approach on 2 846 human whole-body motions and 6 187 natural language descriptions thereof from the KIT Motion-Language Dataset. Our results clearly demonstrate the effectiveness of the proposed model: We show that our model generates a wide variety of realistic motions only from descriptions thereof in form of a single sentence. Conversely, our model is also capable of generating correct and detailed natural language descriptions from human motions. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Human whole-body motion Natural language sequence-to-sequence learning Recurrent neural network

来源：评论

学校读者我要写书评

暂无评论

Question Generation With Doubly Adversarial Nets

引用

IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING 2018年第11期26卷 2230-2239页

作者： Bao, Junwei Gong, Yeyun Duan, Nan Zhou, Ming Zhao, Tiejun Harbin Inst Technol Sch Comp Sci & Technol Harbin 150001 Heilongjiang Peoples R China Microsoft Res Asia Nat Language Comp Grp Beijing 100080 Peoples R China Microsoft Res Asia Beijing 100080 Peoples R China

We study the problem of question generation on a specific domain, where there are no labeled data. To address this problem, we propose a novel neural question generation approach called DoubAN, or doubly adversarial nets, which fully utilizes labeled data from other domains (source domains) and unlabeled data from the target domain. learning a DoubAN involves two adversarial procedures between a question generator and two adversaries. One adversary is a domain-classification discriminator (DC-Dis), which is designed to help the generator learn domain-general representations of the input text. The other is a questionans-wering discriminator (QA-Dis), which provides more training data with estimated reward scores for generated text-question pairs. We conduct experiments on the SQuAD dataset as target-domain unlabeled data and the NewsQA dataset as source-domain labeled data. Experiment results show that our DoubAN achieves better results than baselines. Compared to model variants, which adopt only DC-Dis or QA-Dis, we find that the DC-Dis and QA-Dis indirectly interact with each other and jointly improve the quality of generated questions on the target domain. Moreover, extensive analysis and discussion prove the reasonableness and effectiveness of our proposed approach.

关键词： Question generation doubly-adversarial net sequence-to-sequence learning neural networks natural language processing artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

A Deep Neural Network Model for Joint Entity and Relation Extraction

引用

IEEE ACCESS 2019年 7卷 179143-179150页

作者： Pang, Yihe Liu, Jie Liu, Lizhen Yu, Zhengtao Zhang, Kai Capital Normal Univ Dept Informat Engn Beijing 100048 Peoples R China Capital Normal Univ Res Ctr Language Intelligence Beijing 100048 Peoples R China Beijing Adv Innovat Ctr Imaging Theory & Technol Beijing 100048 Peoples R China Kunming Univ Sci & Technol Dept Informat Engn & Automat Kunming 650504 Yunnan Peoples R China

Joint extraction of entities and their relations from the text is an essential issue in automatic knowledge graph construction, which is also known as the joint extraction of relational triplets. The relational triplets in sentence are complicated, multiple and different relational triplets may have overlaps, which is commonly seen in reality. However, multiple pairs of triplets cannot be efficiently extracted in most of the previous works. To mitigate this problem, we propose a deep neural network model based on the sequenceto-sequence learning, namely, the hybrid dual pointer networks (HDP), which extracts multiple pairs of triplets from the given sentence by generating the hybrid dual pointer sequence. In experiments, we tested our model using the New York Times (NYT) public dataset. The experimental results demonstrated that our model outperformed the state-of-the-art work, and achieved a 17.1% improvement on the F1 values.

关键词： Automatic knowledge graph construction entity and relation extraction deep neural networks natural language processing pointer networks relational triplet extraction sequence-to-sequence learning

来源：评论

学校读者我要写书评

暂无评论

Retrospective Encoders for Video Summarization 15th

Retrospective Encoders for Video Summarization

引用

15th European Conference on Computer Vision (ECCV)

作者： Zhang, Ke Grauman, Kristen Sha, Fei Univ Southern Calif Dept Comp Sci Los Angeles CA 90089 USA Facebook AI Res 300 W Sixth St Austin TX 78701 USA Netflix 5808 Sunset Blvd Los Angeles CA 90028 USA

ISBN: (纸本)9783030012373;9783030012366

Supervised learning techniques have shown substantial progress on video summarization. State-of-the-art approaches mostly regard the predicted summary and the human summary as two sequences (sets), and minimize discriminative losses that measure element-wise discrepancy. Such training objectives do not explicitly model how well the predicted summary preserves semantic information in the video. Moreover, those methods often demand a large amount of human generated summaries. In this paper, we propose a novel sequence-to-sequence learning model to address these deficiencies. The key idea is to complement the discriminative losses with another loss which measures if the predicted summary preserves the same information as in the original video. To this end, we propose to augment standard sequence learning models with an additional "retrospective encoder" that embeds the predicted summary into an abstract semantic space. The embedding is then compared to the embedding of the original video in the same space. The intuition is that both embeddings ought to be close to each other for a video and its corresponding summary. Thus our approach adds to the discriminative loss a metric learning loss that minimizes the distance between such pairs while maximizing the distances between unmatched ones. One important advantage is that the metric learning loss readily allows learning from videos without human generated summaries. Extensive experimental results show that our model outperforms existing ones by a large margin in both supervised and semi-supervised settings.

关键词： Video summarization sequence-to-sequence learning

来源：评论

学校读者我要写书评

暂无评论

RECALL NEURAL NETWORK FOR SOURCE SEPARATION

RECALL NEURAL NETWORK FOR SOURCE SEPARATION

引用

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

作者： Chien, Jen-Tzung Tsou, Kai-Wei Natl Chiao Tung Univ Dept Elect & Comp Engn Hsinchu Taiwan

ISBN: (纸本)9781538646588

This paper presents a novel memory-augmented neural network for single-channel source separation. We propose a recall neural network (RCNN) where a couple of external memories are realized for sequence-to-sequence learning based on an encoder and a decoder. These memories are learned in a two-pass sensing procedure where the mixed signal is encoded and then decoded (or recalled) as context vectors by using a bidirectional long short-term memory (LSTM) and a LSTM, respectively. These context vectors are integrated in a gating layer. A set of attention weights are calculated to attend the hidden state of decoder to implement a recurrent neural network for source separation. A gated attention mechanism is carried out to fulfill a specialized memory network. The regression errors due to two passes of sensing procedure and one pass of gated attention are jointly minimized to estimate the weight parameters of different components in different layers. The experiments on multi-speaker speech enhancement show that the proposed RCNN consistently outperforms LSTM and neural Turing machine in different settings.

关键词： long short-term memory sequence-to-sequence learning recall neural network source separation

来源：评论

学校读者我要写书评

暂无评论

EFFICIENTLY TRAINABLE TEXT-TO-SPEECH SYSTEM BASED ON DEEP CONVOLUTIONAL NETWORKS WITH GUIDED ATTENTION

EFFICIENTLY TRAINABLE TEXT-TO-SPEECH SYSTEM BASED ON DEEP CO...

引用

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

作者： Tachibana, Hideyuki Uenoyama, Katsuya Aihara, Shunsuke PKSHA Technol Inc Bunkyo Ku Tokyo Japan

ISBN: (纸本)9781538646588

This paper describes a novel text-to-speech (TTS) technique based on deep convolutional neural networks (CNN), without use of any recurrent units. Recurrent neural networks (RNN) have become a standard technique to model sequential data recently, and this technique has been used in some cutting-edge neural TTS techniques. However, training RNN components often requires a very powerful computer, or a very long time, typically several days or weeks. Recent other studies, on the other hand, have shown that CNN-based sequence synthesis can be much faster than RNN-based techniques, because of high parallelizability. The objective of this paper is to show that an alternative neural TTS based only on CNN alleviate these economic costs of training. In our experiment, the proposed Deep Convolutional TTS was sufficiently trained overnight (15 hours), using an ordinary gaming PC equipped with two GPUs, while the quality of the synthesized speech was almost acceptable.

关键词： Text-to-speech deep learning convolutional neural network attention sequence-to-sequence learning

来源：评论

学校读者我要写书评

暂无评论

Lemmatization for Ancient Languages: Rules or Neural Networks? 7th

Lemmatization for Ancient Languages: Rules or Neural Network...

引用

7th International Conference on Artificial Intelligence and Natural Language (AINL)

作者： Dereza, Oksana Natl Res Univ Higher Sch Econ Moscow Russia Lomonosov Moscow State Univ Moscow Russia

ISBN: (纸本)9783030012045;9783030012038

Lemmatisation, which is one of the most important stages of text preprocessing, consists in grouping the inflected forms of a word together so they can be analysed as a single item. This task is often considered solved for most modern languages irregardless of their morphological type, but the situation is dramatically different for ancient languages. Rich inflectional system and high level of orthographic variation common to these languages together with lack of resources make lemmatising historical data a challenging task. It becomes more and more important as manuscripts are being extensively digitized now, but still remains poorly covered in literature. In this work, I compare a rule-based and a neural network based approach to lemmatisation in case of Early Irish (Old and Middle Irish are often described together as "Early Irish") data.

关键词： Early Irish Natural language processing Under-resourced languages Lemmatisation Neural networks sequence-to-sequence learning

来源：评论

学校读者我要写书评

暂无评论

RHYTHM-FLEXIBLE VOICE CONVERSION WITHOUT PARALLEL DATA USING CYCLE-GAN OVER PHONEME POSTERIORGRAM sequenceS

RHYTHM-FLEXIBLE VOICE CONVERSION WITHOUT PARALLEL DATA USING...

引用

IEEE Workshop on Spoken Language Technology (SLT)

作者： Yeh, Cheng-chieh Hsu, Po-chun Chou, Ju-chieh Lee, Hung-yi Lee, Lin-shan Natl Taiwan Univ Coll Elect Engn & Comp Sci Taipei Taiwan

ISBN: (纸本)9781538643341

Speaking rate refers to the average number of phonemes within some unit time, while the rhythmic patterns refer to duration distributions for realizations of different phonemes within different phonetic structures. Both are key components of prosody in speech, which is different for different speakers. Models like cycle-consistent adversarial network (Cycle-GAN) and variational auto-encoder (VAE) have been successfully applied to voice conversion tasks without parallel data. However, due to the neural network architectures and feature vectors chosen for these approaches, the length of the predicted utterance has to be fixed to that of the input utterance, which limits the flexibility in mimicking the speaking rates and rhythmic patterns for the target speaker. On the other hand, sequence-to-sequence learning model was used to remove the above length constraint, but parallel training data are needed. In this paper, we propose an approach utilizing sequence-to-sequence model trained with unsupervised Cycle-GAN to perform the transformation between the phoneme posteriorgram sequences for different speakers. In this way, the length constraint mentioned above is removed to offer rhythm-flexible voice conversion without requiring parallel data. Preliminary evaluation on two datasets showed very encouraging results.

关键词： voice conversion sequence-to-sequence learning unsupervised learning cycle-gan

来源：评论

学校读者我要写书评

暂无评论

Logician: A Unified End-to-End Neural Approach for Open-Domain Information Extraction 18

Logician: A Unified End-to-End Neural Approach for Open-Doma...

引用

11th ACM International Conference on Web Search and Data Mining

作者： Sun, Mingming Li, Xu Wang, Xin Fan, Miao Feng, Yue Li, Ping Baidu Res BDL Beijing Peoples R China

ISBN: (纸本)9781450355810

In this paper, we consider the problem of open information extraction (OIE) for extracting entity and relation level intermediate structures from sentences in open-domain. We focus on four types of valuable intermediate structures (Relation, Attribute, Description, and Concept), and propose a unified knowledge expression form, SAOKE, to express them. We publicly release a data set which contains 48,248 sentences and the corresponding facts in the SAOKE format labeled by crowdsourcing. To our knowledge, this is the largest publicly available human labeled data set for open information extraction tasks. Using this labeled SAOKE data set, we train an end-to-end neural model using the sequence-to-sequence paradigm, called Logician, to transform sentences into facts. For each sentence, different to existing algorithms which generally focus on extracting each single fact without concerning other possible facts, Logician performs a global optimization over all possible involved facts, in which facts not only compete with each other to attract the attention of words, but also cooperate to share words. An experimental study on various types of open domain relation extraction tasks reveals the consistent superiority of Logician to other states-of-the-art algorithms. The experiments verify the reasonableness of SAOKE format, the valuableness of SAOKE data set, the effectiveness of the proposed Logician model, and the feasibility of the methodology to apply end-to-end learning paradigm on supervised data sets for the challenging tasks of open information extraction.

关键词： Knowledge expression open information extraction end-to-end learning sequence-to-sequence learning deep learning

来源：评论

学校读者我要写书评

暂无评论

learning explicit video attributes from mid-level representation for video captioning

引用

COMPUTER VISION AND IMAGE UNDERSTANDING 2017年 163卷 126-138页

作者： Nian, Fudong Li, Teng Wang, Yan Wu, Xinyu Ni, Bingbing Xu, Changsheng Anhui Univ Minist Educ Key Lab Intelligent Comp & Signal Proc Hefei Anhui Peoples R China Chinese Acad Sci Inst Automat Natl Lab Pattern Recognit Beijing Peoples R China Chinese Acad Sci Shenzhen Inst Adv Technol Beijing Peoples R China Shanghai Jiao Tong Univ Shanghai Peoples R China

Recent works on video captioning mainly learn the map from low-level visual features to language description directly without explicitly representing the high-level semantic video concepts (e.g. objects, actions in the annotated sentences). To bridge the semantic gap, in this paper, addressing it, we propose a novel video attribute representation learning algorithm for video concept understanding and utilize the learned explicit video attribute representation to improve video captioning performance. To achieve it, firstly, inspired by the success of spectrogram in audio processing, a novel mid-level video representation named "video response map" (VRM) is proposed, by which the frame sequence could be represented by a single image representation. Therefore, the video attributes representation learning could be converted to a well-studied multi-label image classification problem. Then in the captions prediction step, the learned video attributes feature is integrated with the single frame feature to improve previous sequence-to sequence language generation model by adjusting the LSTM (Long-Short Term Memory) input units. The proposed video captioning framework could both handle variable frame inputs and utilize high-level semantic video attribute features. Experimental results on video captioning tasks show that the proposed method, utilizing only RGB frames as input without extra video or text training data, could achieve competitive performance with state-of-the-art methods. Furthermore, the extensive experimental evaluations on the UCF-101 action classification benchmark well demonstrate the representation capability of the proposed VRM. (C) 2017 Elsevier Inc. All rights reserved.

关键词： Mid-level video representation Video attributes learning Video caption sequence-to-sequence 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：