检索结果-内蒙古大学图书馆

6th Chinese Conference on Pattern Recognition (CCPR)

作者： Zhao, Qiyang Li, Weibo Wang, Fan Yin, Baolin Beihang Univ Sch Comp Sci & Engn NLSDE Beijing 100191 Peoples R China Beijing Youth Polit Coll Beijing Peoples R China

ISBN: (纸本)9783662456460;9783662456453

In the state-of-the-art saliency detection methods based on contrast priors, little attention is paid on the region smoothness constraints. The paper proposes a two-stage saliency detection method in which a smoothness prior is explicitly involved in a continuous Conditional Random Field (CRF). In stage one, we construct a continuous CRF based on the sparse codes of perceptual features on all locations, and minimize the energy of CRF to obtain discrimination maps. In stage two, we train a discriminative machine and learn the saliency maps from discrimination maps, aiming to take the human attention priors into consideration. Our experiments on MSRA-1000 show that the new method is effective against the state-of-the-art methods.

关键词： saliency continuous CRF sparse coding

来源：评论

学校读者我要写书评

暂无评论

Iterative sparse coding for Colorization Based Compression 11

Iterative Sparse Coding for Colorization Based Compression

引用

11th International Conference on Image Analysis and Recognition (ICIAR)

作者： Lee, Suk-Ho Oh, Paul Kang, Moon Gi Yonsei Univ Dept Elect & Elect Engn 134 Shinchon Dong Seoul 120749 South Korea Dongseo Univ Dept Software Engn Pusan 617716 South Korea

ISBN: (纸本)9783319117584;9783319117577

Colorization based coding is a technique which compresses a color image using the colorization method. The main issue in colorization based coding is to extract a good RP(representative pixel) set from the original color image from which the colored image can be reconstructed in the decoder to a sufficient level. In this paper, we propose an iterative sparse coding method for the extraction of the RP set. Observations show that the proposed method computes simultaneously the locally optimal RP set and the locally optimal Levin's colorization matrix. Furthermore, experimental results show that the proposed method provides better color image reconstruction and compression rate than conventional colorization based coding methods.

关键词： Colorization Compression Optimization sparse coding Color image

来源：评论

学校读者我要写书评

暂无评论

D-SELD: Dataset-Scalable Exemplar LCA-Decoder

引用

NEUROMORPHIC COMPUTING AND ENGINEERING 2024年第4期4卷 044009-044009页

作者： Takaghaj, Sanaz Mahmoodi Sampson, Jack Penn State Univ Dept Comp Sci & Engn University Pk PA 16802 USA Penn State Univ Sch Engn Design & Innovat University Pk PA 16802 USA

Neuromorphic computing has recently gained significant attention as a promising approach for developing energy-efficient, massively parallel computing systems inspired by the spiking behavior of the human brain and natively mapping spiking neural networks (SNNs). Effective training algorithms for SNNs are imperative for increased adoption of neuromorphic platforms;however, SNN training continues to lag behind advances in other classes of ANN. In this paper, we reduce this gap by proposing an innovative encoder-decoder technique that leverages sparse coding and the locally competitive algorithm (LCA) to provide an algorithm specifically designed for neuromorphic platforms. Using our proposed Dataset-Scalable Exemplar LCA-Decoder we reduce the computational demands and memory requirements associated with training SNNs using error backpropagation methods on increasingly larger training sets. We offer a solution that can be scalably applied to datasets of any size. Our results show the highest reported top-1 test accuracy using SNNs on the ImageNet and CIFAR100 datasets, surpassing previous benchmarks. Specifically, we achieved a record top-1 accuracy of 80.75% on ImageNet (ILSVRC2012 validation set) and 79.32% on CIFAR100 using SNNs.

关键词： encoder-decoder LCA exemplar learning spiking neural networks sparse coding neuromorphic computing

来源：评论

学校读者我要写书评

暂无评论

Fusion sparse coding Algorithm for Impulse Feature Extraction in Machinery Weak Fault Detection

Fusion Sparse Coding Algorithm for Impulse Feature Extractio...

引用

Prognostics and System Health Management Conference (PHM- Hunan)

作者： Deng, Sen Jing, Bo Zhou, Hongliang Air Force Engn Univ Sch Aeronaut & Astronaut Engn Xian 710038 Shaanxi Peoples R China

ISBN: (纸本)9781479979585

Impulse components in vibration signals are important indicators of machinery health states. sparse coding (SC) is regarded as an efficient impulse feature extraction method, but it cannot extract the weak impulse features in vibration signals with heavy background noises. In this paper, a fusion sparse coding (FSC) method is proposed to extract impulse components effectively. Firstly, several sparse coding algorithms are executed in parallel independently as participating algorithms. Then, fusion scheme of different sparse coding algorithms is presented to improve the accuracy of sparse signal reconstruction. Lastly, the proposed method is used to process aircraft engine rotor vibration signals compared with other feature extraction approaches. Experiment result shows FSC method can extract impulse features accurately from heavy noisy vibration signal, and it provides great significance for machinery weak fault detection and diagnosis.

关键词： Impulse feature extraction fault detection sparse coding information fusion

来源：评论

学校读者我要写书评

暂无评论

DYNAMIC sparse coding WITH SMOOTHING PROXIMAL GRADIENT METHOD

DYNAMIC SPARSE CODING WITH SMOOTHING PROXIMAL GRADIENT METHO...

引用

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

作者： Chalasani, Rakesh Principe, Jose C. Univ Florida Dept Elect & Comp Engn Gainesville FL 32611 USA

ISBN: (纸本)9781479928934

In this work we focus on the problem of estimating time-varying sparse signals from a sequence of under-sampled observations. We formulate this problem as estimating hidden states in a dynamic model and exploit the underlying temporal structure to find a more accurate solution, particularly when the information in the observations is at scarce. We propose an optimization procedure based on smoothing proximal gradient method to estimate these hidden states. We show that the proposed model is efficient and more robust to the noise in the system.

关键词： State-space sparse coding Dynamics Proximal methods

来源：评论

学校读者我要写书评

暂无评论

Designing an Efficient Framework for Large-Scale Data Processing and Analysis Based on Deep Learning Technology 24

Designing an Efficient Framework for Large-Scale Data Proces...

引用

3rd International Conference on Cyber Security, Artificial Intelligence and Digital Economy (CSAIDE)

作者： Liu, Qian Wang, Xingda Univ South Australia Adelaide SA Australia Xian Univ Architecture & Technol Xian 710055 Shaanxi Peoples R China

ISBN: (纸本)9798400718212

This paper presents a comprehensive framework designed for the efficient processing and analysis of large-scale data through the integration of deep learning technologies. The framework leverages the capabilities of deep learning to enhance the accuracy and speed of data analysis, addressing the challenges posed by the sheer volume and complexity of modern datasets. It incorporates advanced deep learning models, including convolutional neural networks and sparse coding, to extract and analyze features from large datasets effectively. The framework is built upon a robust architecture that integrates with existing big data technologies like Hadoop, Spark, and TensorFlow, facilitating scalable and distributed data processing. Experimental results demonstrate significant improvements in performance analysis and prediction accuracy across various application domains, highlighting the framework's effectiveness in harnessing deep learning for large-scale data analysis. This study contributes to the field by proposing a scalable, efficient solution for data-driven decision-making and opens new avenues for research in deep learning applications for big data.

关键词： Deep Learning Big Data Data Analysis Convolutional Neural Networks sparse coding

来源：评论

学校读者我要写书评

暂无评论

Characterize and Compare the Performance of Deep Learning Optimizers in Recurrent Neural Network Architectures 48

Characterize and Compare the Performance of Deep Learning Op...

引用

48th Annual IEEE International Computers, Software, and Applications Conference (COMPSAC) - Digital Development for a Better Future

作者： Zaeed, Mohammad Islam, Tanzima Z. Indic, Vladimir Texas State Univ San Marcos TX 78666 USA Univ Novi Sad Novi Sad Serbia

ISBN: (纸本)9798350376975;9798350376968

The emergence of newer hardware and limited accessibility to that hardware are driving the need to characterize application performance on commonly used architectures to gain valuable insights. While abundant performance events create opportunities for detailed performance characterization, they also make the process overwhelming. This paper takes a systematic approach to address the problem of performance characterization by taking advantage of such opportunities. Specifically, this paper 1) characterizes the hardware usage behaviors of commonly used optimizers in the context of Recurrent Neural Network (RNN) architectures, 2) connects observations to identify problems and opportunities for optimization in those optimizers, and 3) creates a large performance dataset. Comparing the hardware resource usage behaviors of Gradient Descent (GS), AdaGrad (AdaGrad), and ADAM (ADAM) optimizers for three commonly used recurrent neural network-based deep learning architectures running on AMD MI100 GPUs, this paper identifies non-uniform memory access as one of the most frequent issues and suggests strategies for optimizing such problems.

关键词： Recurrent Neural Network High-Performance Computing Performance Analytics sparse coding Optimization

来源：评论

学校读者我要写书评

暂无评论

ADAPTIVELY HIERARCHICAL QUANTIZATION VARIATIONAL AUTOENCODER BASED ON FEATURE DECOUPLING AND SEMANTIC CONSISTENCY FOR IMAGE GENERATION 31

ADAPTIVELY HIERARCHICAL QUANTIZATION VARIATIONAL AUTOENCODER...

引用

2024 International Conference on Image Processing

作者： Zhang, Ying Park, Hyunhee Jia, Hanchao Wang, Fan Samsung R&D Inst China Beijing Beijing Peoples R China Samsung Elect Dept Camera Innovat Grp Suwon South Korea

ISBN: (纸本)9798350349405;9798350349399

The Vector Quantized Variational AutoEncoder (VQ-VAE) has shown great potential in image generation, especially the methods with hierarchical features. However, the lack of decoupling of structural information between hierarchical features leads to semantic inconsistencies and redundant structural features, resulting in incompatible outputs. In this study, we propose the Adaptively Hierarchical Quantization Variational AutoEncoder (AHQ-VAE) to generate high-fidelity images with a unified structure. To ensure the semantic consistency of continuous space, we employ the Spatially Consistent Semantic Embedding (SCSE) module to align the hierarchical features, while decoupling global structural information and local details. To ensure the consistency of discrete space, we introduce the Adaptive Bottom Quantizer (ABQ) to generate the quantized bottom codes consistent with quantized top codes, so that the local details can adapt to the global semantics. Extensive experiments demonstrate our approach can generate high-quality images with a unified structure.

关键词： Image generation VQ-VAE Feature decoupling Hierarchical consistency sparse coding

来源：评论

学校读者我要写书评

暂无评论

Deep sparse Representation Classification for Aero-engine Inter-shaft Bearing Fault Diagnosis

Deep Sparse Representation Classification for Aero-engine In...

引用

IEEE International Conference on Prognostics and Health Management (ICPHM)

作者： Yao, Renhe Jiang, Hongkai Liu, Yunpeng Wang, Xin Shao, Haidong Jiang, Wenxin Northwestern Polytech Univ Sch Civil Aviat Xian Peoples R China Hunan Univ Coll Mech & Vehicle Engn Changsha Peoples R China

ISBN: (纸本)9798350374483;9798350374476

Fault diagnosis of aero-engine inter-shaft bearing under variable operating conditions poses a significant challenge in the industry. Existing sparse classification methods with shallow architectures suffer from insufficient fault feature extraction and interference removal capabilities with limited training samples, resulting in low diagnostic accuracies. To address this issue, this study introduces an approach termed deep sparse representation classification (DSRC). DSRC seamlessly integrates multiple layers for dictionary learning and sparse coding. In the initial phase, the dictionary learning layer is employed to acquire the Fisher discriminative sparse representation information, while the sparse coding layer is utilized to eliminate interfering components and simultaneously enhance sparsity. The incorporation of a weight matrix, guided by a high-energy atom selection strategy, links the upward and downward processes of dictionary learning and sparse coding. Subsequently, the frequency-weighted energy operator kurtosis-based feature vectors are extracted from the reconstructed signals of the newly acquired dictionary and coding coefficients. Ultimately, these discriminative feature vectors are directly input into a straightforward classifier for intelligent fault diagnosis. DSRC is applied to an aero-engine inter-shaft bearing fault data under multiple speeds. Results demonstrate that it can effectively realize discriminative fault feature extraction and high-precision automatic fault identification.

关键词： Aero-engine inter-shaft bearing Fault diagnosis sparse classification Dictionary learning sparse coding

来源：评论

学校读者我要写书评

暂无评论

A PRECONDITIONING APPROACH TO OPTIMIZING SENSING MATRIX FOR IMPROVED COMPRESSED SENSING CT RECONSTRUCTION 31

A PRECONDITIONING APPROACH TO OPTIMIZING SENSING MATRIX FOR ...

引用

2024 International Conference on Image Processing

作者： Theeda, Prasad Ting, Chee-Ming Pal, Arghya Ombao, Hernando Monash Univ Sch Informat Technol Malaysia Campus Subang Jaya Malaysia King Abdullah Univ Sci & Technol Stat Program Thuwal Saudi Arabia Vellore Inst Technol VIT Vellore Tamil Nadu India

ISBN: (纸本)9798350349405;9798350349399

Compressed sensing (CS) exploiting inherent sparsity prior of signals has been proven effective for sparse-view computed tomography (CT) image reconstruction from undersampled projection data. However, most CS-based CT studies focused on formulating different sparsity regularizers, e.g., total variation (TV) minimization, and neglect design of an incoherent sensing matrix - a key factor of CS performance. The sensing matrix formed by an incomplete set of Radon projections in CT typically exhibits large coherence. In this paper, we propose a novel method for optimizing the sensing matrix via preconditioning to improve CS-CT reconstruction. A well-conditioned preconditioner is designed to optimally reduce the coherence of the sensing matrix and thus improving the CS systems. The desired preconditioner is obtained by solving a nonconvex optimization problem via gradient descent method. The preconditioned systems solved by TV-based sparse recovery algorithms can provide better reconstruction accuracy with fewer measurements even in noisy settings. Evaluated on brain and COVID-19 chest CT datasets, the proposed method when used for preconditioning of Radon sensing matrix reconstructed images with substantially higher quality with faster speed than baselines without preconditioning.

关键词： CT reconstruction compressed sensing sparse coding preconditioning TVAL3 ADMM

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：