检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Zhang, Wenlong Liu, Yihao Dong, Chao Qiao, Yu ShenZhen Key Lab of Computer Vision and Pattern Recognition SIAT-SenseTime Joint Lab. Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China University of Chinese Academy of Sciences

Generative Adversarial Networks (GAN) have demonstrated the potential to recover realistic details for single image super-resolution (SISR). To further improve the visual quality of super-resolved results, PIRM2018-SR Challenge employed perceptual metrics to assess the perceptual quality, such as PI, NIQE, and Ma. However, existing methods cannot directly optimize these indifferentiable perceptual metrics, which are shown to be highly correlated with human ratings. To address the problem, we propose Super-Resolution Generative Adversarial Networks with Ranker (RankSRGAN) to optimize generator in the direction of perceptual metrics. Specifically, we first train a Ranker which can learn the behavior of perceptual metrics and then introduce a novel rank-content loss to optimize the perceptual quality. The most appealing part is that the proposed method can combine the strengths of different SR methods to generate better results. Extensive experiments show that RankSRGAN achieves visually pleasing results and reaches state-of-the-art performance in perceptual metrics. Project page: https://***/Projects/RankSRGAN. Copyright © 2019, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Gaussian meta-embeddings for efficient scoring of a heavy-tailed PLDA model

Gaussian meta-embeddings for efficient scoring of a heavy-ta...

引用

2018 Speaker and Language Recognition Workshop, ODYSSEY 2018

作者： Brümmer, Niko Silnova, Anna Burget, Lukáš Stafylakis, Themos Nuance Communications South Africa Brno University of Technology Czech Republic Computer Vision Lab. University of Nottingham Omilia Conversational Intelligence United Kingdom

Embeddings in machine learning are low-dimensional representations of complex input patterns, with the property that simple geometric operations like Euclidean distances and dot products can be used for classification and comparison tasks. We introduce meta-embeddings, which live in more general inner product spaces and which are designed to better propagate uncertainty through the embedding bottleneck. Traditional embeddings are trained to maximize between-class and minimize within-class distances. Meta-embeddings are trained to maximize relevant information throughput. As a proof of concept in speaker recognition, we derive an extractor from the familiar generative Gaussian PLDA model (GPLDA). We show that GPLDA likelihood ratio scores are given by Hilbert space inner products between Gaussian likelihood functions, which we term Gaussian meta-embeddings (GMEs). Meta-embedding extractors can be generatively or discriminatively trained. GMEs extracted by GPLDA have fixed precisions and do not propagate uncertainty. We show that a generalization to heavy-tailed PLDA gives GMEs with variable precisions, which do propagate uncertainty. Experiments on NIST SRE 2010 and 2016 show that the proposed method applied to i-vectors without length normalization is up to 20% more accurate than GPLDA applied to length-normalized i-vectors. © ODYSSEY 2018. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

DeepMine Speech Processing Database: Text-Dependent and Independent Speaker Verification and Speech Recognition in Persian and English

DeepMine Speech Processing Database: Text-Dependent and Inde...

引用

2018 Speaker and Language Recognition Workshop, ODYSSEY 2018

作者： Zeinali, Hossein Sameti, Hossein Stafylakis, Themos Sharif DeepMine Ltd. Tehran Iran Sharif University of Technology Tehran Iran Computer Vision Lab. University of Nottingham Omilia Conversational Intelligence United Kingdom

In this paper, we introduce a new database for text-dependent, text-prompted and text-independent speaker recognition, as well as for speech recognition. DeepMine is a large-scale database in Persian and English, with its current version containing more than 1300 speakers and 360 thousand recordings overall. DeepMine has several appealing characteristics which make it unique of its kind. First of all, it is the first large-scale speaker recognition database in Persian, enabling the development of voice biometrics applications in the native language of about 110 million people. Second, it is the largest text-dependent and text-prompted speaker recognition database in English, facilitating research on deep learning and other data demanding approaches. Third, its unique combination of Persian and English makes it suitable for exploring domain adaptation and transfer learning approaches, which constitute some of the emerging tasks in speech and speaker recognition. Finally, the extensive annotation with respect to age, gender, province, and educational level, combined with the inherent variability of the Persian language in terms of different accents are ideal for exploring the use of attribute information in utterance and speaker modeling. The presentation of the database is accompanied with several experiments using state-of-the-art algorithms. More specifically, we conduct experiments using HMM-based i-vectors, and we reaffirm their effectiveness in text-dependent speaker recognition. Furthermore, we conduct speech recognition experiments using the annotated text-independent part of the database for training and testing, and we demonstrate that the database can also serve for training robust speech recognition models in Persian. © ODYSSEY 2018. All rights reserved.

关键词： Database systems

来源：评论

学校读者我要写书评

暂无评论

Sensations on the body: Varying duration, overlap, ratios and placement on dynamic vibrotactile patterns 18

Sensations on the body: Varying duration, overlap, ratios an...

引用

19th International Conference on Human-computer Interaction, Interaccion 2018

作者： Morrison, Ann Manresa-Yee, Cristina Jensen, Walther Digital Life Lab. University of Southern Queensland Toowoomba Australia Computer Science and Mathematics Science Department University of Balearic Islands Palma Spain Department of Architecture Design and Media Technology Aalborg University Denmark

ISBN: (纸本)9781450364911

Prior research revealed that people have strong responses to various vibrotactile patterns acting on the trunk of the body. Moreover, they experienced patterns on the back and around the waist more positively compared to those on the front or mid-front of the body. In this study, we use a vibrotactile vest designed according to Traditional Chinese principles to identify the effect of individual factors in the patterns. We examined variations of duration, overlap, placement and repetition. Our research questions are: 'What influenced peoples' perception of the sensations?' 'How can we design for the wearer with specific sensations in mind?' As an example, we found using actuators with the same placement but varying the ratio between duration and overlap gave wearers' opposing perceived effects. One ratio was perceived as calming/feel good (0.5), another as activating (>0.5), another again as annoying ( © 2018 Association for Computing Machinery.

关键词： Wearable computers

来源：评论

学校读者我要写书评

暂无评论

A review of uncertainty quantification in deep learning: Techniques, applications and challenges

arXiv

引用

arXiv 2020年

作者： Abdar, Moloud Pourpanah, Farhad Hussain, Sadiq Rezazadegan, Dana Liu, Li Ghavamzadeh, Mohammad Fieguth, Paul Cao, Xiaochun Khosravi, Abbas Rajendra Acharya, U. Makarenkov, Vladimir Nahavandi, Saeid Deakin University Australia College of Mathematics and Statistics Guangdong Key Lab. of Intelligent Information Processing Shenzhen University Shenzhen518060 China Dibrugarh University Dibrugarh India Department of Computer Science and Software Engineering Swinburne University of Technology Melbourne Australia Center for Machine Vision and Signal Analysis University of Oulu Oulu Finland Google research United States Department of Systems Design Engineering University of Waterloo Waterloo Canada State Key Laboratory of Information Security Institute of Information Engineering Chinese Academy of Sciences Beijing China Department of Electronics and Computer Engineering Ngee Ann Polytechnic Clementi Singapore Department of Computer Science University of Quebec in Montreal MontrealQC Canada

—Uncertainty quantification (UQ) plays a pivotal role in the reduction of uncertainties during both optimization and decision making, applied to solve a variety of real-world applications in science and engineering. Bayesian approximation and ensemble learning techniques are two of the most widely-used UQ methods in the literature. In this regard, researchers have proposed different UQ methods and examined their performance in a variety of applications such as computer vision (e.g., self-driving cars and object detection), image processing (e.g., image restoration), medical image analysis (e.g., medical image classification and segmentation), natural language processing (e.g., text classification, social media texts and recidivism risk-scoring), bioinformatics, etc. This study reviews recent advances in UQ methods used in deep learning, investigates the application of these methods in reinforcement learning, and highlight the fundamental research challenges and directions associated with the UQ field. Copyright © 2020, The Authors. All rights reserved.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Author Correction: Predicting standardized uptake value of brown adipose tissue from CT scans using convolutional neural networks

引用

Nature communications 2024年第1期15卷 10022页

作者： Ertunc Erdil Anton S Becker Moritz Schwyzer Borja Martinez-Tellez Jonatan R Ruiz Thomas Sartoretti H Alberto Vargas A Irene Burger Alin Chirindel Damian Wild Nicola Zamboni Bart Deplancke Vincent Gardeux Claudia Irene Maushart Matthias Johannes Betz Christian Wolfrum Ender Konukoglu Computer Vision Lab. ETH Zurich Zurich Switzerland. ertunc.erdil@vision.ee.ethz.ch. Computer Vision Lab. ETH Zurich Zurich Switzerland. Department of Radiology Memorial Sloan Kettering Cancer Center New York NY USA. Institute for Diagnostic and Interventional Radiology University Hospital Zurich Zurich Switzerland. Department of Radiology NYU Grossman School of Medicine New York NY USA. Department of Nursing Physiotherapy and Medicine and SPORT Research Group (CTS-1024) CERNEP Research Center University of Almería Almería Spain. CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN) Instituto de Salud Carlos III Granada Spain. Department of Medicine Division of Endocrinology and Einthoven Laboratory for Experimental Vascular Medicine Leiden University Medical Center Leiden The Netherlands. Department of Physical Education and Sports Faculty of Sports Science Sport and Health University Research Institute (iMUDS) University of Granada 18071 Granada Spain. Instituto de Investigación Biosanitaria Ibs.Granada Granada Spain. Centro de Investigación Biomédica en Red Fisiopatología de la Obesidad y Nutrición (CIBERobn) Instituto de Salud Carlos III Madrid Spain. Department of Nuclear Medicine University Zurich Hospital Zurich Switzerland. University of Zurich Zurich Switzerland. Department of Radiology and Nuclear Medicine University Hospital of Basel Basel Switzerland. Swiss Multi-Omics Center ETH Zürich Zürich Switzerland. Laboratory of Systems Biology and Genetics Institute of Bioengineering School of Life Sciences Ecole Polytechnique Fédérale de Lausanne (EPFL) 1015 Lausanne Switzerland. Swiss Institute of Bioinformatics 1015 Lausanne Switzerland. Department of Endocrinology Diabetes and Metabolism University Hospital Basel and University of Basel Basel Switzerland. Department of Health Sciences and Technology ETH Zurich Zurich Switzerland. The LOOP Zürich - Medical Research Center Zürich Switzerland.

来源：评论

学校读者我要写书评

暂无评论

Underwater photogrammetry in very shallow waters: Main challenges and caustics effect removal

Underwater photogrammetry in very shallow waters: Main chall...

引用

2018 ISPRS TC II Mid-term Symposium "Towards Photogrammetry 2020"

作者： Agrafiotis, P. Skarlatos, D. Forbes, T. Poullis, C. Skamantzari, M. Georgopoulos, A. Cyprus University of Technology Civil Engineering and Geomatics Dept. Lab of Photogrammetric Vision 2-8 Saripolou str. Limassol3036 Cyprus Immersive and Creative Technologies Lab. Department of Computer Science and Software Engineering Concordia University QuebecH3G 1M8 Canada National Technical University of Athens School of Rural and Surveying Engineering Lab. of Photogrammetry Zografou Campus 9 Heroon Polytechniou str. Zografou Athens15780 Greece

In this paper, main challenges of underwater photogrammetry in shallow waters are described and analysed. The very short camera to object distance in such cases, as well as buoyancy issues, wave effects and turbidity of the waters are challenges to be resolved. Additionally, the major challenge of all, caustics, is addressed by a new approach for caustics removal (Forbes et al., 2018) which is applied in order to investigate its performance in terms of SfM-MVS and 3D reconstruction results. In the proposed approach the complex problem of removing caustics effects is addressed by classifying and then removing them from the images. We propose and test a novel solution based on two small and easily trainable Convolutional Neural Networks (CNNs). Real ground truth for caustics is not easily availab.e. We show how a small set of synthetic data can be used to train the network and later transfer the le arning to real data with robustness to intra-class variation. The proposed solution results in caustic-free images which can be further used for other tasks as may be needed. © Authors 2018.

关键词： Underwater photogrammetry

来源：评论

学校读者我要写书评

暂无评论

Anonymization of lab.led TOF-MRA images for brain vessel segmentation using generative adversarial networks

arXiv

引用

arXiv 2020年

作者： Kossen, Tabea Subramaniam, Pooja Madai, Vince I. Hennemuth, Anja Hildebrand, Kristian Hilbert, Adam Sobesky, Jan Livne, Michelle Galinovic, Ivana Khalil, Ahmed A. Fiebach, Jochen B. Frey, Dietmar CLAIM - Charité Lab for AI in Medicine Charité Universitätsmedizin Berlin Germany Department of Computer Engineering and Microelectronics Computer Vision & Remote Sensing Technical University Berlin Berlin Germany Department of Electrical Engineering and Computer Science Technical University of Berlin Berlin Germany School of Computing and Digital Technology Faculty of Computing Engineering and the Built Environment Birmingham City University Birmingham United Kingdom Institute for Imaging Science and Computational Modelling in Cardiovascular Medicine Charité Universitätsmedizin Berlin Berlin Germany Fraunhofer MEVIS Bremen Germany Department VI Computer Science and Media Beuth University of Applied Sciences Berlin Germany Johanna-Etienne-Hospital Neuss Germany Centre for Stroke Research Berlin Charité Universitätsmedizin Berlin Berlin Germany Department of Neurology Max Planck Institute for Human Cognitive and Brain Sciences Leipzig Germany Mind Brain Body Institute Berlin School of Mind and Brain Humboldt University Berlin Berlin Germany Berlin Institute of Health Berlin Germany

Anonymization and data sharing are crucial for privacy protection and acquisition of large datasets for robust medical image analysis. This represents a major challenge, especially for brain imaging research. Here, the unique structure of brain images allows for potential re-identification and thus requires anonymization beyond conventional methods. Generative adversarial networks (GANs) have the potential to provide anonymous images while maintaining their predictive properties. Analyzing brain vessel segmentation as a use case, we trained 3 GAN architectures on time-of-flight (TOF) magnetic resonance angiography (MRA) patches of patients with cerebrovascular disease for image-lab.l pair generation: 1) Deep convolutional GAN, 2) Wasserstein-GAN with gradient penalty (WGAN-GP) and 3) WGAN-GP with spectral normalization (WGAN-GP-SN). First, the synthesized image-lab.ls from each GAN architecture were used to train a U-net for vessel segmentation. The U-nets were then tested on real patient data. In total, 66 patients were used for this analysis. In a second step, we simulated the application of our synthetic patches in a transfer learning approach using a second, independent dataset. Here, for an increasing number of up to 15 patients we evaluated vessel segmentation model performance on real data with and without pre-training on generated patches. Finally, performance for all models was assessed by the Dice Similarity Coefficient (DSC) and the 95th percentile of the Hausdorff Distance (95HD). Comparing the 3 GAN architectures, the U-net model trained on synthetic data generated by the WGAN-GP-SN showed the highest performance to predict brain vessels (DSC/95HD 0.82/28.97) benchmarked by the U-net trained on real data (0.89/26.61). The transfer learning approach showed superior performance for the same GAN architecture compared to no pre-training, especially for one lab.led patient only (DSC/95HD 0.91/25.68 compared to DSC/95HD 0.85/27.36). In a brain imaging segment

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Smart Cloud System for Forensic Thermal Image Enhancement Using Local and Global Logarithmic Transform Histogram Matching

Smart Cloud System for Forensic Thermal Image Enhancement Us...

引用

IEEE International Conference on Smart Cloud (SmartCloud)

作者： Viacheslav Voronin Evgenii Semenishchev Vladimir Frants Sos Agaian Lab. “Mathematical methods of image processing and intelligent computer vision systems” Don State Technical University Rostov-on-Don Russian Federation “STANKIN” Moscow State University of Technology Moscow Russian Federation Dept. of Computer Science CUNY/The College of Staten Island Staten Island New York United States

Digital images used in the investigation of a crime often undergo several concurrent enhancement operations for improved automated analysis. The challenges are related to the big size of data and complexity of the forensic image processing. Our purpose is providing a smart cloud system to image processing for PC and Smartphones with limited computation complexity. This paper presents a new thermal image enhancement algorithm based on combined local and global image processing in the frequency domain. The presented approach uses the fact that the relationship between stimulus and perception is logarithmic. The basic idea is to apply logarithmic transform histogram matching with spatial equalization approach on different image blocks. The resulting image is a weighted mean of all processing blocks. The weights for every local and global enhanced image driven through optimization of measure of enhancement (EME). Some presented experimental results illustrate the performance of the proposed cloud system on real thermal images in comparison with the traditional methods.

关键词： Histograms Transforms Image enhancement Forensics Adaptive equalizers Clouds

来源：评论

学校读者我要写书评

暂无评论

Blind Visual Quality Assessment for Smart Cloud-Based Video Storage

Blind Visual Quality Assessment for Smart Cloud-Based Video ...

引用

IEEE International Conference on Smart Cloud (SmartCloud)

作者： Vladimir Frants Viacheslav Voronin Alexander Zelenskiy Sos Agaian Metrological lab. “Small GIC” Moscow State University of Technology “STANKIN” Moscow Russian Federation Pro-rector for Research Work and R&D Politics Moscow State University of Technology “STANKIN” Moscow Russian Federation Lab. “Mathematical methods of image processing and intelligent computer vision systems.” Don State Technical University Rostov-on-Don Russian Federation Dept. of Computer Science CUNY The College of Staten Island Staten Island New York United States

In this paper, we present a new video quality metric targeted for use within cloud-based video storage systems. Because of the limited capacity of storage solutions currently in use, it is common for stored videos to have low perceived video quality. Ability to predict the quality of coded video sequence in fully automatic fashion is crucial for optimal coding parameters selection. We have developed general quality estimation approach applicable for different kinds of video content and useful for online correction of perceived quality of stored video sequence. Proposed objective video quality metric is designed to have a high level of correlation with human-based quality assessment results. A high level of conformity of predicted video quality to perceived quality level is achieved by using a convolutional neural network trained on a large volume of video data in the framework of generative adversarial learning and combination of carefully selected regularization techniques. Evaluation of developed VQA metric on commonly used datasets shows equal or better correlation with MOS than the current state of the art approaches.

关键词： Quality assessment Measurement Video recording Video sequences Streaming media Training Visualization

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：