检索结果-内蒙古大学图书馆

Inpainting with Separable Mask Update Convolution network

SENSORS 2023年第15期23卷 6689-6689页

作者： Gong, Jun Luo, Senlin Yu, Wenxin Nie, Liang Beijing Inst Technol Informat Syst & Secur & Countermeasures Expt Ctr Beijing 100081 Peoples R China Southwest Univ Sci & Technol Sch Comp Sci & Technol Mianyang 621010 Peoples R China

Image inpainting is an active area of research in image processing that focuses on reconstructing damaged or missing parts of an image. The advent of deep learning has greatly advanced the field of image restoration in recent years. While there are many existing methods that can produce high-quality restoration results, they often struggle when dealing with images that have large missing areas, resulting in blurry and artifact-filled outcomes. This is primarily because of the presence of invalid information in the inpainting region, which interferes with the inpainting process. To tackle this challenge, the paper proposes a novel approach called separable mask update convolution. This technique automatically learns and updates the mask, which represents the missing area, to better control the influence of invalid information within the mask area on the restoration results. Furthermore, this convolution method reduces the number of network parameters and the size of the model. The paper also introduces a regional normalization technique that collaborates with separable mask update convolution layers for improved feature extraction, thereby enhancing the quality of the restored image. Experimental results demonstrate that the proposed method performs well in restoring images with large missing areas and outperforms state-of-the-art image inpainting methods significantly in terms of image quality.

关键词： image inpainting image processing separable mask update convolution encoder-decoder network

来源：评论

学校读者我要写书评

暂无评论

Motion-Based Wave Inference With Neural networks: Transfer Learning From Numerical Simulation to Experimental Data

引用

JOURNAL OF OFFSHORE MECHANICS AND ARCTIC ENGINEERING-TRANSACTIONS OF THE ASME 2024年第5期146卷 051204-051204页

作者： Bisinotto, Gustavo A. de Mello, Pedro C. Cozman, Fabio G. Tannuri, Eduardo A. Univ Sao Paulo Numer OffshoreTank TPN BR-05508030 Sao Paulo SP Brazil Univ Sao Paulo BR-05508030 Sao Paulo SP Brazil

The directional wave spectrum, which describes the distribution of wave energy along frequencies and directions, can be estimated from the measured motions of a vessel subjected to a particular sea condition by resorting to the wave-buoy analogy. Several methods have been proposed to address the inverse estimation problem;recently, machine learning techniques have been assessed as further alternatives. However, it may be difficult to gather large datasets of in-service motion responses and the associated sea states to train effective data-driven models. In this work, an encoder-decoder neural network is trained with the synthetic responses of a station-keeping platform supply vessel (PSV) to estimate the directional wave spectrum. This estimation model is directly applied to perform wave inference from motion data of wave basin tests with a small-scale model of the same vessel. Furthermore, fine-tuning is also used to incorporate experimental data into the neural network model. Results show a satisfactory match between estimated and measured values, both with respect to the energy distribution and the integral spectrum parameters, indicating that the proposed approach can be employed to obtain data-driven wave inference models when there is little or no availability of measured motion records and the corresponding sea conditions.

关键词： directional wave spectrum encoder-decoder network transfer learning wave-buoy analogy platform supply vessel model tests ocean waves and associated statistics

来源：评论

学校读者我要写书评

暂无评论

Unabridged adjacent modulation for clothing parsing

引用

PATTERN RECOGNITION 2022年 127卷 108594-108594页

作者： Zhang, Dong Zuo, Chengting Wu, Qianhao Fu, Liyong Xiang, Xinguang Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing 210094 Peoples R China Chinese Acad Forestry Res Inst Forest Resource Informat Tech Beijing 10 0 091 Beijing 100091 Peoples R China Natl Forestry & Grassland Adm Key Lab Forest Management & Growth Modeling Beijing 100091 Peoples R China

Clothing parsing has made tremendous progress in the domain of computer vision recently. Most state-ofthe-art methods are based on the encoder-decoder architecture. However, the existing methods mainly neglect problems of feature uncalibration within blocks and semantics dilution between blocks. In this work, we propose an unabridged adjacent modulation network (UAM-Net) to aggregate multi-level features for clothing parsing. We first build an unabridged channel attention (UCA) mechanism on feature maps within each block for feature recalibration. We further design a top-down adjacent modulation (TAM) for decoder blocks. By deploying TAM, high-level semantic information and visual contexts can be gradually transferred into lower-level layers without loss. The joint implementation of UCA and TAM ensures that the encoder has an enhanced feature representation ability, and the low-level features of the decoders contain abundant semantic contexts. Quantitative and qualitative experimental results on two challenging benchmarks (i.e., colorful fashion parsing and the modified fashion clothing) declare that our proposed UAM-Net can achieve competitive high-accurate performance with the state-of-the-art methods. The source codes are available at: https://***/ctzuo/UAM-Net .(c) 2022 Elsevier Ltd. All rights reserved.

关键词： encoder-decoder network Clothing parsing Attention learning Features modulation Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Robust acoustic scene classification using a multi-spectrogram encoder-decoder framework

引用

DIGITAL SIGNAL PROCESSING 2021年 110卷 102943-102943页

作者： Pham, Lam Phan, Huy Nguyen, Truc Palaniappan, Ramaswamy Mertins, Alfred McLoughlin, Ian Univ Kent Sch Comp Medway Kent England Queen Mary Univ London Sch Elect Engn & Comp Sci London England Graz Univ Technol Signal Proc & Speech Commun Lab Graz Austria Univ Lubeck Inst Signal Proc Lubeck Germany Singapore Inst Technol Singapore Singapore

This article proposes an encoder-decoder network model for Acoustic Scene Classification (ASC), the task of identifying the scene of an audio recording from its acoustic signature. We make use of multiple low-level spectrogram features at the front-end, transformed into higher level features through a welltrained CNN-DNNfront-end encoder. The high-level features and their combination (via a trained feature combiner) are then fed into different decoder models comprising random forest regression, DNNs and a mixture of experts, for back-end classification. We conduct extensive experiments to evaluate the performance of this framework on various ASC datasets, including LITIS Rouen and IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events (DCASE) 2016 Task 1, 2017 Task 1, 2018 Tasks 1A & 1B and 2019 Tasks 1A & 1B. The experimental results highlight two main contributions;the first is an effective method for high-level feature extraction from multi-spectrogram input via the novel CNN-DNNarchitecture encoder network, and the second is the proposed decoder which enables the framework to achieve competitive results on various datasets. The fact that a single framework is highly competitive for several different challenges is an indicator of its robustness for performing general ASC tasks. (C) 2020 Elsevier Inc. All rights reserved.

关键词： Acoustic scene classification encoder-decoder network Low-level features High-level features Multi-spectrogram

来源：评论

学校读者我要写书评

暂无评论

DRNet: Segmentation and localization of optic disc and Fovea from diabetic retinopathy image

引用

ARTIFICIAL INTELLIGENCE IN MEDICINE 2021年 111卷 102001-102001页

作者： Hasan, Md. Kamrul Alam, Md. Ashraful Elahi, Md. Toufick E. Roy, Shidhartho Marti, Robert Khulna Univ Engn Technol Dept Elect & Elect Engn Khulna 9203 Bangladesh Univ Girona Comp Vis & Robot Inst Girona Spain

Background and objective: In modern ophthalmology, automated Computer-aided Screening Tools (CSTs) are crucial non-intrusive diagnosis methods, where an accurate segmentation of Optic Disc (OD) and localization of OD and Fovea centers are substantial integral parts. However, designing such an automated tool remains challenging due to small dataset sizes, inconsistency in spatial, texture, and shape information of the OD and Fovea, and the presence of different artifacts. Methods: This article proposes an end-to-end encoder-decoder network, named DRNet, for the segmentation and localization of OD and Fovea centers. In our DRNet, we propose a skip connection, named residual skip connection, for compensating the spatial information lost due to pooling in the encoder. Unlike the earlier skip connection in the UNet, the proposed skip connection does not directly concatenate low-level feature maps from the encoder's beginning layers with the corresponding same scale decoder. We validate DRNet using different publicly available datasets, such as IDRiD, RIMONE, DRISHTI-GS, and DRIVE for OD segmentation;IDRiD and HRF for OD center localization;and IDRiD for Fovea center localization. Results: The proposed DRNet, for OD segmentation, achieves mean Intersection over Union (mIoU) of 0.845, 0.901, 0.933, and 0.920 for IDRiD, RIMONE, DRISHTI-GS, and DRIVE, respectively. Our OD segmentation result, in terms of mIoU, outperforms the state-of-the-art results for IDRiD and DRIVE datasets, whereas it outperforms state-of-the-art results concerning mean sensitivity for RIMONE and DRISHTI-GS datasets. The DRNet localizes the OD center with mean Euclidean Distance (mED) of 20.23 and 13.34 pixels, respectively, for IDRiD and HRF datasets;it outperforms the state-of-the-art by 4.62 pixels for IDRiD dataset. The DRNet also successfully localizes the Fovea center with mED of 41.87 pixels for the IDRiD dataset, outperforming the state-of-the-art by 1.59 pixels for the same dataset. Conclusion

关键词： Diabetic retinopathy and glaucoma Ophthalmology encoder-decoder network Skip connection Segmentation and localization

来源：评论

学校读者我要写书评

暂无评论

Crack segmentation through deep convolutional neural networks and heterogeneous image fusion

引用

AUTOMATION IN CONSTRUCTION 2021年 125卷 103605-103605页

作者： Zhou, Shanglian Song, Wei Univ Alabama Dept Civil Construct & Environm Engn Tuscaloosa AL 35487 USA

A DCNN-based crack segmentation methodology is proposed by leveraging heterogeneous image fusion to alleviate image-related disturbances in intensity or range image data and mitigate uncertainties through crossdomain (i.e., intensity and range data domains) feature correlation. Intensity and range images are captured from concrete roadways and integrated through data fusion. Three encoder-decoder networks representing different patterns on exploiting the image data (i.e., fused raw image, raw range image, filtered range image, and raw intensity image) are proposed and compared to benchmarks. Experimental results demonstrate the proposed DCNN exploiting the fused raw image through an ?extract-fuse? pattern achieves the most robust and accurate performance on crack segmentation among the implemented DCNNs.

关键词： Concrete roadway encoder-decoder network Heterogeneous image fusion Intensity data Range data Crack segmentation

来源：评论

学校读者我要写书评

暂无评论

A 3D Reconstruction Framework of Buildings Using Single Off-Nadir Satellite Image

引用

REMOTE SENSING 2021年第21期13卷 4434-4434页

作者： Zhao, Chunhui Zhang, Chi Yan, Yiming Su, Nan Harbin Engn Univ Coll Informat & Commun Engn Harbin 150001 Peoples R China Harbin Engn Univ Minist Ind & Informat Technol Key Lab Adv Marine Commun & Informat Technol Harbin 150001 Peoples R China

A novel framework for 3D reconstruction of buildings based on a single off-nadir satellite image is proposed in this paper. Compared with the traditional methods of reconstruction using multiple images in remote sensing, recovering 3D information that utilizes the single image can reduce the demands of reconstruction tasks from the perspective of input data. It solves the problem that multiple images suitable for traditional reconstruction methods cannot be acquired in some regions, where remote sensing resources are scarce. However, it is difficult to reconstruct a 3D model containing a complete shape and accurate scale from a single image. The geometric constraints are not sufficient as the view-angle, size of buildings, and spatial resolution of images are different among remote sensing images. To solve this problem, the reconstruction framework proposed consists of two convolutional neural networks: Scale-Occupancy-network (Scale-ONet) and model scale optimization network (Optim-Net). Through reconstruction using the single off-nadir satellite image, Scale-Onet can generate water-tight mesh models with the exact shape and rough scale of buildings. Meanwhile, the Optim-Net can reduce the error of scale for these mesh models. Finally, the complete reconstructed scene is recovered by Model-Image matching. Profiting from well-designed networks, our framework has good robustness for different input images, with different view-angle, size of buildings, and spatial resolution. Experimental results show that an ideal reconstruction accuracy can be obtained both on the model shape and scale of buildings.

关键词： 3D model reconstruction building satellite image encoder-decoder network

来源：评论

学校读者我要写书评

暂无评论

Dual-stream pyramid registration network

引用

MEDICAL IMAGE ANALYSIS 2022年第0期78卷 1页

作者： Kang, Miao Hu, Xiaojun Huang, Weilin Scott, Matthew R. Reyes, Mauricio Malong LLC Wilmington DE 19890 USA Xi An Jiao Tong Univ Inst Artificial Intelligence & Robot Xian Peoples R China Univ Bern ARTORG Ctr Biomed Engn Res Bern Switzerland

We propose a Dual-stream Pyramid Registration network (referred as Dual-PRNet) for unsupervised 3D brain image registration. Unlike recent CNN-based registration approaches, such as VoxelMorph, which computes a registration field from a pair of 3D volumes using a single-stream network, we design a two-stream architecture able to estimate multi-level registration fields sequentially from a pair of feature pyramids. Our main contributions are: (i) we design a two-stream 3D encoder-decoder network that computes two convolutional feature pyramids separately from two input volumes;(ii) we propose sequential pyramid registration where a sequence of pyramid registration (PR) modules is designed to predict multi-level registration fields directly from the decoding feature pyramids. The registration fields are refined gradually in a coarse-to-fine manner via sequential warping, which equips the model with a strong capability for handling large deformations;(iii) the PR modules can be further enhanced by computing local 3D correlations between the feature pyramids, resulting in the improved Dual-PRNet(++) able to aggregate rich detailed anatomical structure of the brain;(iv) our Dual-PRNet(++) can be integrated into a 3D segmentation framework for joint registration and segmentation, by precisely warping voxellevel annotations. Our methods are evaluated on two standard benchmarks for brain MRI registration, where Dual-PRNet(++) outperforms the state-of-the-art approaches by a large margin, i.e., improving recent VoxelMorph from 0.511 to 0.748 (Dice score) on the Mindboggle101 dataset. In addition, we further demonstrate that our methods can greatly facilitate the segmentation task in a joint learning framework, by leveraging limited annotations. (c) 2022PublishedbyElsevierB.V.

关键词： Medical image registration encoder-decoder network Deformable registration 3D segmentation Brain MRI

来源：评论

学校读者我要写书评

暂无评论

MIANet: Multi-level temporal information aggregation in mixed-periodicity time series forecasting tasks

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2023年第1期121卷

作者： Wang, Sheng Chen, Xi Ma, Dongliang Wang, Chen Wang, Yong Qi, Honggang Zhou, Gongjian Li, Qingli Liu, Min East China Normal Univ Key Lab Geog Informat Sci Minist Educ China Shanghai Peoples R China East China Normal Univ Sch Geog Sci Shanghai Peoples R China Minist Nat Resources Key Lab Spatial temp Big Data Anal & Applicat Nat Shanghai Peoples R China East China Normal Univ Shanghai Key Lab Multidimens Informat Proc Shanghai Peoples R China Anhui Univ Sch Resources & Environm Engn Hefei Peoples R China Sun Yat sen Univ Sch Aeronaut & Astronaut Guangzhou Peoples R China Univ Chinese Acad Sci Sch Comp Sci & Technol Beijing Peoples R China Harbin Inst Technol Sch Elect & Informat Engn Harbin Peoples R China

Regular human activities generate a large number of time series with mixed periodicity that can reflect human behavior patterns and the societal working mechanism. When forecasting these time series, nonlinear neural networks often encounter some limitations, such as utilizing mixed-periodic patterns, balancing multi-level information, incorporating future vision, forecasting delays and scale insensitivity, which affect the forecasting accuracy. To address these problems, we propose the Multi-level Information Aggregation network (MIANet), a novel neural network with four key characteristics: (i) a novel folded recurrent structure that dynamically updates the local and mini-local information at a global range in a compact manner;(ii) a new recurrent unit called Folded Convolution Aggregation Temporal Memory (FCATM) that extracts and aggregates neighbor-trends in local and mini-local data;(iii) a fusing decoder structure that promotes the sharing of forward-backward future information and adaptively adjusts relationships among adjacent points;and (iv) a new Skip-Autoregressive (SAR) linear strategy that addresses scale sensitivity issues. The SAR can be embedded as a plug-and-play component into other deep learning (DL) models. Compared with other baseline methods, MIANet obtains statistically significant improvements on six real-world datasets, as demonstrated by conducting two-sample t-tests, indicating that the MIANet can be applied to various predictive scenarios, such as road occupancy, electricity consumption, pedestrian flow and urban noise.

关键词： Deep learning Data mining Time series forecasting Sequence modeling encoder-decoder network Time series with mixed periodicity

来源：评论

学校读者我要写书评

暂无评论

Swin-MFINet: Swin transformer based multi-feature integration network for detection of pixel-level surface defects

引用

EXPERT SYSTEMS WITH APPLICATIONS 2022年 209卷

作者： Uzen, Huseyin Turkoglu, Muammer Yanikoglu, Berrin Hanbay, Davut Bingo Univ Dept Comp Engn TR-12000 Bingo Turkey Samsun Univ Dept Software Engn TR-55000 Samsun Turkey Sabanci Univ Fac Engn & Nat Sci TR-34956 Istanbul Turkey Inonu Univ Dept Comp Engn TR-44000 Malatya Turkey

Automatic surface defect detection is critical for manufacturing industries, such as steel, fabric, and marble industries. This study proposes a Swin transformer-based model called Multi-Feature Integration network (Swin-MFINet) for pixel-level surface defect detection. The proposed model consists of an encoder, a Swin transformer-based decoder, and Multi-Feature Integration (MFI) modules. In the encoder module of the proposed model, a pre-trained Inception network is used to extract key features from small-size datasets. In the decoder section, global semantic features are obtained from the initial features by using the Swin-transformer block, which is the newest transformer technology of today. In addition, the convolution layer is used in the last step of the decoder, since transformers are limited in acquiring small spatial details such as edges, colors, and textures, which are important in detecting some small defects. In the last module called MFI, feature maps from different decoder stages are combined, and the channel squeeze-spatial excitation block is applied to reveal important features. Finally, a prediction map is obtained by applying a convolution layer and sigmoid activation function to the MFI module output, respectively. The performance of proposed model is analyzed over MT and MVTec datasets containing surface defect images. The proposed model obtained mIoU scores of 81.37%, and 77.07% respectively, for these two datasets These results outperform the state-of-the-art for the surface defect detection problem.

关键词： Pixel-Level Surface Defects Detection Swin Transformers encoder-decoder network Convolutional Neural network

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：