检索结果-内蒙古大学图书馆

Unsupervised Representation Disentanglement Using Cross Domain Features and Adversarial Learning in variational autoencoder Based Voice Conversion

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2020年第4期4卷 468-479页

作者： Huang, Wen-Chin Luo, Hao Hwang, Hsin-Te Lo, Chen-Chou Peng, Yu-Huai Tsao, Yu Wang, Hsin-Min Acad Sinica Inst Informat Sci Taipei 11529 Taiwan Nagoya Univ Grad Sch Informat Nagoya Aichi 4648601 Japan Acad Sinica Res Ctr Informat Technol Inst Informat Sci Taipei 11529 Taiwan

An effective approach for voice conversion (VC) is to disentangle linguistic content from other components in the speech signal. The effectiveness of variational autoencoder (VAE) based VC (VAE-VC), for instance, strongly relies on this principle. In our prior work, we proposed a cross-domain VAE-VC (CDVAE-VC) framework, which utilized acoustic features of different properties, to improve the performance of VAE-VC. We believed that the success came from more disentangled latent representations. In this article, we extend the CDVAE-VC framework by incorporating the concept of adversarial learning, in order to further increase the degree of disentanglement, thereby improving the quality and similarity of converted speech. More specifically, we first investigate the effectiveness of incorporating the generative adversarial networks (GANs) with CDVAE-VC. Then, we consider the concept of domain adversarial training and acid an explicit constraint to the latent representation, realized by a speaker classifier, to explicitly eliminate the speaker information that resides in the latent code. Experimental results confirm that the degree of disentanglement of the learned latent representation can he enhanced by both GANs and the speaker classifier. Meanwhile, subjective evaluation results in terms of quality and similarity scores demonstrate the effectiveness of our proposed methods.

关键词： Voice conversion unsupervised learning disentangled representation variational autoencoder adversarial learning cross domain features

来源：评论

学校读者我要写书评

暂无评论

A variational autoencoder Mixture Model for Online Behavior Recommendation

引用

IEEE ACCESS 2020年 8卷 132736-132747页

作者： Nguyen, Minh-Duc Cho, Yoon-Sik Sejong Univ Dept Software Convergence Seoul 05006 South Korea Sejong Univ Dept Data Sci Seoul 05006 South Korea

Online behavior recommendation is an attractive research topic related to social media mining. This topic focuses on suggesting suitable behaviors for users in online platforms, including music listening, video watching, e-commerce, to name but a few to improve the user experience, an essential factor for the success of online services. A successful online behavior recommendation system should have the ability to predict behaviors that users used to performs and also suggest behaviors that users never performed before. In this paper, we develop a mixture model that contains two components to address this problem. The first component is the user-specific preference component that represents the habits of users based on their behavior history. The second component is the latent group preference component based on variational autoencoder, a deep generative neural network. This component corresponds to the hidden interests of users and allows us to discover the unseen behavior of users. We conduct experiments on various real-world datasets with different characteristics to show the performance of our model in different situations. The result indicates that our proposed model outperforms the previous mixture models for recommendation problem.

关键词： Mixture models History Probabilistic logic Data models Neural networks Task analysis Training Online behavior recommendation mixture model variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Semi-Supervised Neural Chord Estimation Based on a variational autoencoder With Latent Chord Labels and Features

引用

IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING 2020年 28卷 2956-2966页

作者： Wu, Yiming Carsault, Tristan Nakamura, Eita Yoshii, Kazuyoshi Kyoto Univ Grad Sch Informat Kyoto 6068501 Japan IRCAM Representat Mus F-75004 Paris France Kyoto Univ Hakubi Ctr Adv Res Kyoto 6068501 Japan

This paper describes a statistically-principled semi-supervised method of automatic chord estimation (ACE) that can make effective use of music signals regardless of the availability of chord annotations. The typical approach to ACE is to train a deep classification model (neural chord estimator) in a supervised manner by using only annotated music signals. In this discriminative approach, prior knowledge about chord label sequences (model output) has scarcely been taken into account. In contrast, we propose a unified generative and discriminative approach in the framework of amortized variational inference. More specifically, we formulate a deep generative model that represents the generative process of chroma vectors (observed variables) from discrete labels and continuous features (latent variables), which are assumed to follow a Markov model favoring self-transitions and a standard Gaussian distribution, respectively. Given chroma vectors as observed data, the posterior distributions of the latent labels and features are computed approximately by using deep classification and recognition models, respectively. These three models form a variational autoencoder and can be trained jointly in a semi-supervised manner. The experimental results show that the regularization of the classification model based on the Markov prior of chord labels and the generative model of chroma vectors improved the performance of ACE even under the supervised condition. The semi-supervised learning using additional non-annotated data can further improve the performance.

关键词： Hidden Markov models Music Multiple signal classification Markov processes Computational modeling Estimation Automatic chord estimation semi-supervised learning variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Remote sensing image captioning via variational autoencoder and Reinforcement Learning

引用

KNOWLEDGE-BASED SYSTEMS 2020年 203卷 105920-105920页

作者： Shen, Xiangqing Liu, Bing Zhou, Yong Zhao, Jiaqi Liu, Mingming China Univ Min & Technol Sch Comp Sci & Technol Xuzhou 221116 Jiangsu Peoples R China Minist Educ Mine Digitizat Engn Res Ctr Beijing Peoples R China Chinese Acad Sci Inst Elect Beijing 100190 Peoples R China Jiangsu Vocat Inst Architectural Technol Sch Intelligent Mfg Xuzhou 221008 Jiangsu Peoples R China Jiangsu Normal Univ Sch Mechatron Engn Xuzhou 221008 Jiangsu Peoples R China

Image captioning, i.e., generating the natural semantic descriptions of given image, is an essential task for machines to understand the content of the image. Remote sensing image captioning is a part of the field. Most of the current remote sensing image captioning models suffered the overfitting problem and failed to utilize the semantic information in images. To this end, we propose a variational autoencoder and Reinforcement Learning based Two-stage Multi-task Learning Model (VRTMM) for the remote sensing image captioning task. In the first stage, we finetune the CNN jointly with the variational autoencoder. In the second stage, the Transformer generates the text description using both spatial and semantic features. Reinforcement Learning is then applied to enhance the quality of the generated sentences. Our model surpasses the previous state of the art records by a large margin on all seven scores on Remote Sensing Image Caption Dataset. The experiment result indicates our model is effective on remote sensing image captioning and achieves the new state-of-the-art result. (C) 2020 Elsevier B.V. All rights reserved.

关键词： Transformer variational autoencoder Transfer learning Remote sensing image captioning Self-attention mechanisms Convolutional neural network Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

adVAE: A self-adversarial variational autoencoder with Gaussian anomaly prior knowledge for anomaly detection

引用

KNOWLEDGE-BASED SYSTEMS 2020年 190卷 105187-105187页

作者： Wang, Xuhong Du, Ying Lin, Shijie Cui, Ping Shen, Yuntian Yang, Yupu Shanghai Jiao Tong Univ Shanghai Peoples R China Wuhan Univ Wuhan Peoples R China Univ Calif Davis Davis CA 95616 USA

Recently, deep generative models have become increasingly popular in unsupervised anomaly detection. However, deep generative models aim at recovering the data distribution rather than detecting anomalies. Moreover, deep generative models have the risk of overfitting training samples, which has disastrous effects on anomaly detection performance. To solve the above two problems, we propose a self-adversarial variational autoencoder (adVAE) with a Gaussian anomaly prior assumption. We assume that both the anomalous and the normal prior distribution are Gaussian and have overlaps in the latent space. Therefore, a Gaussian transformer net T is trained to synthesize anomalous but near-normal latent variables. Keeping the original training objective of a variational autoencoder, a generator G tries to distinguish between the normal latent variables encoded by E and the anomalous latent variables synthesized by T, and the encoder E is trained to discriminate whether the output of G is real. These new objectives we added not only give both G and E the ability to discriminate, but also become an additional regularization mechanism to prevent overfitting. Compared with other competitive methods, the proposed model achieves significant improvements in extensive experiments. The employed datasets and our model are available in a Github repository. (C) 2019 Elsevier B.V. All rights reserved.

关键词： Anomaly detection Outlier detection Novelty detection Deep generative model variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

variational autoencoder Anomaly-Detection of Avalanche Deposits in Satellite SAR Imagery 2020

Variational Autoencoder Anomaly-Detection of Avalanche Depos...

引用

Proceedings of the 10th International Conference on Climate Informatics

作者： Saumya Sinha Sophie Giffard-Roisin Fatima Karbou Michael Deschatres Anna Karas Nicolas Eckert Cécile Coléou Claire Monteleoni University of Colorado Boulder USA ISTerre Univ. Grenoble Alpes Univ. Savoie Mont Blanc CNRS IRD IFSTTAR France CNRM-GAME Météo-France Grenoble France Irstea Université Grenoble Alpes France Irstea Université Grenoble Alpes France

ISBN: (纸本)9781450388481

This work demonstrates that deep unsupervised learning holds much promise for rare event detection, even when labeled data is limited. Rare event detection is challenging for traditional supervised learning approaches due to high class-imbalance. We demonstrate the efficacy of deep unsupervised learning, in particular a variational autoencoder (VAE), when used for anomaly detection, in an application to avalanche detection in the French Alps, from satellite SAR imagery and a limited on-the-ground survey. Remarkably, our results demonstrate that supervision (i.e., access to labeled data) is only needed in the validation phase of our training pipeline, to tune a hyper-parameter: the threshold on reconstruction error of the VAE (trained in an unsupervised fashion) that will be used to designate an observation as an anomaly, i.e., an avalanche in this context. Our method outperforms previous benchmarks in avalanche detection, including supervised learning a convolutional neural network on an artificially balanced version of the same data. To the best of our knowledge, this is the first step in exploring the potential of applying deep unsupervised learning methods to detect avalanche deposits. Our proposed semi-supervised learning pipeline is also promising for real-world settings in which the data is very unbalanced and labels are limited.

关键词： unsupervised deep learning variational autoencoder anomaly detection avalanche detection satellite imaging

来源：评论

学校读者我要写书评

暂无评论

Dimension reduction and 2D-visualization for early change of state detection in a machining process with a variational autoencoder approach

引用

INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY 2020年第11-12期111卷 3597-3611页

作者： Proteau, Antoine Zemouri, Ryad Tahan, Antoine Thomas, Marc Ecole Technol Super Dept Mech Engn Montreal PQ H3C 1K3 Canada HESAM Univ CEDRIC Lab Conservatoire Natl Arts & Metiers CNAM F-750141 Paris France

In this paper, we applied a variational autoencoder approach to an industrial machining problematic. We proposed a model based on a two-steps training process and a two-dimensional latent space. This two-dimensional latent space has better dimension reduction capability compared to a principal component analysis, which would require 24 components to express 90.0% of the variation. Moreover, the proposed model is shown capable of classifying a cutting operation based solely on data obtained from sensors mounted on a CNC machine, with an accuracy of 99.24%. The suggested model is also shown capable to be an efficient visual process monitoring tool capable of detecting early changes of state in a machining process. We show that this approach can visually identify defect caused by an increase of less than 1% of the energy in the signal, which is earlier than conventional monitoring methods. Additionally, our work is based on an industrial dataset acquired during regular production. This increases the opportunity for technological transfer when it comes to better understanding and better monitoring early changes in a machining process.

关键词： variational autoencoder Dimension reduction 2D-visualization Process monitoring Early detection Machining process

来源：评论

学校读者我要写书评

暂无评论

GBNF-VAE: A Pathological Voice Enhancement Model Based on Gold Section for Bottleneck Feature With variational autoencoder

引用

Journal of Voice 2023年

作者： Liu, Ganjun Zhang, Tao Ding, Biyun Lv, Ying Hou, Xiaohui Guo, Haoyang Wu, Yaqin Fu, Dehui School of Electrical and Information Engineering Tianjin University Tianjin China School of Software Shanxi Agricultural University Shanxi China Department of Otolaryngology Tianjin Medical University Tianjin China

Objective: Speech enhancement has become a promising technique to accommodate demands of the improvement in quality of a degraded speech signal. The main works now focus on separating normal speech from noise, but have neglected the low quality of impaired speech influenced by anomalous glottis flow. In order to effectively enhance the pathological speech, it is essential to design a separation mechanism for extracting high-dimensional timbre features and speech features separately to suppress low-dimensional noises. Methods: In this paper, we propose an enhancement model GBNF-VAE to extract timbre efficiently by reducing anomalous airflow noise interference, and by combining the semantic features with timbre features to synthesize the enhanced speech. In particular, the bottleneck feature can characterize the timbre by the controlled number of nodes through the Golden Section method, which effectively improves computational efficiency. In addition, variational autoencoder is adopted to extract semantic features which are combined with the previous timbre features to synthesize the enhanced speech. Results: Finally, spectrum observation, objective indicators and subjective evaluation all show the outstanding performance of GBNF-VAE in pathological speech quality enhancement. © 2023 The Voice Foundation

关键词： Bottleneck feature Golden section Pathological speech enhancement variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Condition-Transforming variational autoencoder for Generating Diverse Short Text Conversations

引用

ACM TRANSACTIONS ON ASIAN AND LOW-RESOURCE LANGUAGE INFORMATION PROCESSING 2020年第6期19卷 1–13页

作者： Ruan, Yu-Ping Ling, Zhen-Hua Zhu, Xiaodan Univ Sci & Technol China Natl Engn Lab Speech & Language Informat Proc 443 Huangshan Rd Hefei Peoples R China Queens Univ Dept Elect & Comp Engn Walter Light HallUnion St Kingston ON K7L 2N1 Canada

In this article, conditional-transforming variational autoencoders (CTVAEs) are proposed for generating diverse short text conversations. In conditional variational autoencoders (CVAEs), the prior distribution of latent variable z follows a multivariate Gaussian distribution with mean and variance modulated by the input conditions. Previous work found that this distribution tended to become condition-independent in practical applications. Thus, this article designs CTVARs to enhance the influence of conditions in CVAEs. In a CTVAE model, the latent variable z is sampled by performing a non-linear transformation on the combination of the input conditions and the samples from a condition-independent prior distribution N(0.1). In our experiments using a Chinese Sina Weibo dataset, the CTVAE model derives z samples for decoding with better condition-dependency than that of the CVAE model. The earth mover's distance (EMD) between the distributions of the latent variable z at the training stage, and the testing stage is also reduced by using the CTVAE model. In subjective preference tests, our proposed CTVAE model performs significantly better than CVAE and sequence-to-sequence (Seq2Seq) models on generating diverse, informative, and topic-relevant responses.

关键词： Neural network variational autoencoder conversation text generation

来源：评论

学校读者我要写书评

暂无评论

variational autoencoder: An unsupervised model for encoding and decoding fMRI activity in visual cortex

引用

NEUROIMAGE 2019年 198卷 125-136页

作者： Han, Kuan Wen, Haiguang Shi, Junxing Lu, Kun-Han Zhang, Yizhen Fu, Di Liu, Zhongming Weldon Sch Biomed Engn W Lafayette IN USA Sch Elect & Comp Engn W Lafayette IN USA Purdue Univ Purdue Inst Integrat Neurosci W Lafayette IN 47906 USA

Goal-driven and feedforward-only convolutional neural networks (CNN) have been shown to be able to predict and decode cortical responses to natural images or videos. Here, we explored an alternative deep neural network, variational auto-encoder (VAE), as a computational model of the visual cortex. We trained a VAE with a five-layer encoder and a five-layer decoder to learn visual representations from a diverse set of unlabeled images. Using the trained VAE, we predicted and decoded cortical activity observed with functional magnetic resonance imaging (fMRI) from three human subjects passively watching natural videos. Compared to CNN, VAE could predict the video-evoked cortical responses with comparable accuracy in early visual areas, but relatively lower accuracy in higher-order visual areas. The distinction between CNN and VAE in terms of encoding performance was primarily attributed to their different learning objectives, rather than their different model architecture or number of parameters. Despite lower encoding accuracies, VAE offered a more convenient strategy for decoding the fMRI activity to reconstruct the video input, by first converting the fMRI activity to the VAE's latent variables, and then converting the latent variables to the reconstructed video frames through the VAE's decoder. This strategy was more advantageous than alternative decoding methods, e.g. partial least squares regression, for being able to reconstruct both the spatial structure and color of the visual input. Such findings highlight VAE as an unsupervised model for learning visual representation, as well as its potential and limitations for explaining cortical responses and reconstructing naturalistic and diverse visual experiences.

关键词： Neural coding variational autoencoder Bayesian brain Visual reconstruction

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：