检索结果-内蒙古大学图书馆

Underwater image restoration using deep encoder-decoder network with symmetric skip connections

SIGNAL IMAGE AND VIDEO PROCESSING 2022年第1期16卷 247-255页

作者： Gangisetty, Shankar Rai, Raghu Raj KLE Technol Univ Hubballi India

Underwater images get degraded for a variety of naturally occurring attributes like haze, suspended particles, light scattering and water types. The primary cause of the degradation is underwater light attenuation that varies with wavelength, unlike the uniform attenuation that occurs in-air. In this paper, we propose an end-to-end deep convolutional neural network architecture to restore the underwater images and improve their visual perception. The encoder learns to encode the degraded image to a lower-dimensional feature map, while the decoder learns to restore the image to a degradation-free form. This is achieved due to the utilization of symmetric skip connections between the encoder-decoder blocks for the propagation of feature maps to improve the sharpness of the restored image and prevent the loss of details caused by the convolutions. We exhaustively evaluate the performance of our network both qualitatively and quantitatively on standard datasets, and the effectiveness of our network is demonstrated with existing methods of underwater image restoration and enhancement techniques.

关键词： Underwater image restoration and enhancement Convolutional neural networks Skip connections encoder-decoder UIEB dataset

来源：评论

学校读者我要写书评

暂无评论

STEDNet: Swin transformer-based encoder-decoder network for noise reduction in low-dose CT

引用

MEDICAL PHYSICS 2023年第7期50卷 4443-4458页

作者： Zhu, Linlin Han, Yu Xi, Xiaoqi Fu, Huijuan Tan, Siyu Liu, Mengnan Yang, Shuangzhan Liu, Chang Li, Lei Yan, Bin PLA Strateg Support Force Informat Engn Univ Henan Key Lab Imaging & Intelligent Proc Zhengzhou Peoples R China Zhengzhou Univ Sch Cyber Sci & Engn Zhengzhou Peoples R China PLA Strateg Support Force Informat Engn Univ Henan Key Lab Imaging & Intelligent Proc Zhengzhou 450001 Peoples R China

BackgroundLow-dose computed tomography (LDCT) can reduce the dose of X-ray radiation, making it increasingly significant for routine clinical diagnosis and treatment planning. However, the noise introduced by low-dose X-ray exposure degrades the quality of CT images, affecting the accuracy of clinical *** noises, artifacts, and high-frequency components are similarly distributed in LDCT images. Transformer can capture global context information in an attentional manner to create distant dependencies on targets and extract more powerful features. In this paper, we reduce the impact of image errors on the ability to retain detailed information and improve the noise suppression performance by fully mining the distribution characteristics of image information. Purpose MethodsThis paper proposed an LDCT noise and artifact suppressing network based on Swin Transformer. The network includes a noise extraction sub-network and a noise removal sub-network. The noise extraction and removal capability are improved using a coarse extraction network of high-frequency features based on full convolution. The noise removal sub-network improves the network's ability to extract relevant image features by using a Swin Transformer with a shift window as an encoder-decoder and skip connections for global feature fusion. Also, the perceptual field is extended by extracting multi-scale features of the images to recover the spatial resolution of the feature maps. The network uses a loss constraint with a combination of L1 and MS-SSIM to improve and ensure the stability and denoising effect of the network. ResultsThe denoising ability and clinical applicability of the methods were tested using clinical datasets. Compared with DnCNN, RED-CNN, CBDNet and TSCN, the STEDNet method shows a better denoising effect on RMSE and PSNR. The STEDNet method effectively removes image noise and preserves the image structure to the maximum extent, making the reconstructed image closest to the NDCT

关键词： deep learning encoder-decoder low-dose CT Swin transformer

来源：评论

学校读者我要写书评

暂无评论

encoder-decoder with Atrous Separable Convolution for Semantic Image Segmentation 1

引用

15th European Conference on Computer Vision (ECCV)

作者： Chen, Liang-Chieh Zhu, Yukun Papandreou, George Schroff, Florian Adam, Hartwig Google Inc Mountain View CA 94043 USA

ISBN: (数字)9783030012342

ISBN: (纸本)9783030012342;9783030012335

Spatial pyramid pooling module or encode-decoder structure are used in deep neural networks for semantic segmentation task. The former networks are able to encode multi-scale contextual information by probing the incoming features with filters or pooling operations at multiple rates and multiple effective fields-of-view, while the latter networks can capture sharper object boundaries by gradually recovering the spatial information. In this work, we propose to combine the advantages from both methods. Specifically, our proposed model, DeepLabv3+, extends DeepLabv3 by adding a simple yet effective decoder module to refine the segmentation results especially along object boundaries. We further explore the Xception model and apply the depthwise separable convolution to both Atrous Spatial Pyramid Pooling and decoder modules, resulting in a faster and stronger encoder-decoder network. We demonstrate the effectiveness of the proposed model on PASCAL VOC 2012 and Cityscapes datasets, achieving the test set performance of 89% and 82.1% without any post-processing. Our paper is accompanied with a publicly available reference implementation of the proposed models in Tensorflow at https://***/tensorflow/models/tree/master/research/deeplab.

关键词： Semantic image segmentation Spatial pyramid pooling encoder-decoder Depthwise separable convolution

来源：评论

学校读者我要写书评

暂无评论

Short-term ship roll motion prediction using the encoder-decoder Bi-LSTM with teacher forcing

引用

OCEAN ENGINEERING 2024年 295卷

作者： Li, Shiyang Wang, Tongtong Li, Guoyuan Skulstad, Robert Zhang, Houxiang Norwegian Univ Sci & Technol Dept Ocean Operat & Civil Engn Alesund 6009 Norway

The safety of maritime operations has become a paramount concern with the advancement of intelligent ships. Ship stability and safety are directly impacted by roll motion, making the prediction of short-term ship roll motion pivotal for assisting navigators in making timely adjustments and averting hazardous roll conditions. However, predicting ship roll motion poses challenges due to nonlinear dynamics. This study aims to predict short-term ship roll motion by leveraging the encoder-decoder structure of Bidirectional Long Short -Term Memory Networks (Bi-LSTM) with teacher forcing. The model is accomplished by employing an encoder- decoder structure to map input sequences to output sequences of varying lengths, and employing teacher forcing to enhance the network's ability to extract information. To refine and analyze the prediction model, aspects such as the quantity of training data to guarantee model generalization, establishing apposite length relationships between input and output sequences, and assessing the model performance in various sea states are investigated. Additionally, comparative experiments assessing roll motion prediction for intervals of 10s, 30 s, 60 s, and 120 s are conducted to substantiate the necessity and effectiveness of the proposed network. The dataset originates from a commercial professional simulator developed by the Norwegian company Offshore Simulator Center AS (OSC).

关键词： Ship roll motion prediction encoder-decoder Bi-LSTM Teacher forcing

来源：评论

学校读者我要写书评

暂无评论

DC-EDN: densely connected encoder-decoder network with reinforced depthwise convolution for face alignment

引用

APPLIED INTELLIGENCE 2021年第7期51卷 5025-5039页

作者： Yang, Lianping Zhang, Hongliang Wei, Panpan Sun, Yubo Zhang, Xiangde Northeastern Univ Coll Sci Shenyang 110819 Liaoning Peoples R China Zhengzhou Univ Sch Elect Engn Zhengzhou Peoples R China

High accuracy and fast face alignment algorithms play an important role in many face-related applications. Generally, the model speed is inversely related to the number of parameters. We construct our network based on densely connected encoder-decoders, which is an efficient method to balance the parameter number and localization results. In each encoder-decoder, we introduce stacking depthwise convolution and depthwise feature fusion within the same channel, which greatly improves the performance of depthwise convolution and reduces the number of model parameters. In addition, we enhance the mean square loss function by assigning different penalty weights to each coordinate according to the distance to the position corresponding to the maximum value in the label heatmap. Experiments show that the model with the improved loss function obtains better localization results. In the experiment, we compare our method to state-of-the-art methods based on 300W and WFLW. The localization error is 2.76% with the common subset of 300W and the model size (0.7M) is small and even utilizes approximately 1% of the number of parameters of the other models. The dataset and model based on WFLW are publicly available at https://***/iam-zhanghongliang/DC-EDN.

关键词： Face alignment Stacking depthwise convolution Depthwise feature fusion encoder-decoder Heatmap

来源：评论

学校读者我要写书评

暂无评论

Dual attention guided deep encoder-decoder network for change analysis in land use/land cover for Dakshina Kannada District, Karnataka, India

引用

ENVIRONMENTAL EARTH SCIENCES 2023年第1期82卷 1-21页

作者： Naik, Nitesh Chandrasekaran, Kandasamy Sundaram, Venkatesan Meenakshi Panneer, Prabhavathy Natl Inst Technol Dept Comp Sci & Engn Mangalore 575025 Karnataka India Natl Inst Technol Dept Comp Sci & Engn Karaikal 609609 Puducherry India Vellore Insitute Technol Sch Informat Technol & Engn Vellore 632014 Tamil Nadu India

The Earth is frequently changed by natural occurrences and human actions that have threatened our environment to a certain extent. Therefore, accurate and timely monitoring of transformations at the surface of the Earth is crucial for precisely facing their harmful effects and consequences. This paper aims to perform a change detection (CD) analysis and assessment of the Dakshina Kannada region, being one of the coastal districts of Karnataka, India. The spatial and temporal variations in land use and land cover (LULC) are being monitored and examined from the data received as LULC maps from the National Remote Sensing Agency, Indian Space Research Organization, India. The time-series data from advanced wide-field sensor (AWiFS) Resourcesat2 satellite as LULC maps (1:250k) are analyzed using a deep learning approach with an encoder-decoder architecture with dual-attention modules for the change analysis. The model provides an overall accuracy and meanIOU(intersection over union) of 94.11% and 74.1%. The LULC maps from 2005 to 2018 (13 years) are utilized to decide the variations in the LULC, including urban development, agricultural variations, vegetation dynamics, forest areas, barren land, littoral swamp, and water bodies, current fallow, etc. The multiclass area-wise changes in terms of percentage show a decline in most LULC classes, which raises a point of concern for the environmental safety of the considered area, which is highly exposed to coastal flooding due to increased urbanization.

关键词： Land use and land cover encoder-decoder Deep learning Dual-attention

来源：评论

学校读者我要写书评

暂无评论

Multi-scale deep encoder-decoder network for salient object detection

引用

NEUROCOMPUTING 2018年 316卷 95-104页

作者： Ren, Qinghua Hu, Renjie Southeast Univ Sch Elect Engn Nanjing Jiangsu Peoples R China

Deep convolutional neural networks (CNNs) have recently made revolutionary improvements in salient object detection. However, most existing CNN-based models fail to precisely separate the whole salient object(s) from a cluttered background due to the downsampling effects or the patch-level operation. In this paper, we propose a multi-scale deep encoder-decoder network which learns discriminative saliency cues and computes confidence scores in an end-to-end fashion. The encoder network extracts meaningful and informative features in a global view, and the decoder network recovers lost detailed object structure in a local perspective. By taking multiple resized images as the inputs, the proposed model incorporates multi-scale features from a shared network and predicts a fine-grained saliency map at the pixel level. To easily and efficiently train the whole network, the light-weighted decoder breaks through the limit of conventional symmetric structure. In addition, a two-stage training strategy is designed to encourage the robustness and accuracy of the network. Without any post-processing steps, our method is capable of significantly reducing the computation complexity while densely segmenting foreground objects from an image. Extensive experiments on six challenging datasets demonstrate that the proposed model outperforms other state-of-the-art approaches in terms of various evaluation metrics. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Salient object detection Multi-scale encoder-decoder CNN

来源：评论

学校读者我要写书评

暂无评论

Deep Hierarchical encoder-decoder Network for Image Captioning

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2019年第11期21卷 2942-2956页

作者： Xiao, Xinyu Wang, Lingfeng Ding, Kun Xiang, Shiming Pan, Chunhong Chinese Acad Sci Inst Automat Natl Lab Pattern Recognit Beijing 100190 Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing Peoples R China

encoder-decoder models have been widely used in image captioning, and most of them are designed via single long short term memory (LSTM). The capacity of single-layer network, whose encoder and decoder are integrated together, is limited for such a complex task of image captioning. Moreover, how to effectively increase the "vertical depth" of encoder-decoder remains to be solved. To deal with these problems, a novel deep hierarchical encoder-decoder network is proposed for image captioning, where a deep hierarchical structure is explored to separate the functions of encoder and decoder. This model is capable of efficiently exerting the representation capacity of deep networks to fuse high level semantics of vision and language in generating captions. Specifically, visual representations in top levels of abstraction are simultaneously considered, and each of these levels is associated to one LSTM. The bottom-most LSTM is applied as the encoder of textual inputs. The application of the middle layer in encoder-decoder is to enhance the decoding ability of top-most LSTM. Furthermore, depending on the introduction of semantic enhancement module of image feature and distribution combine module of text feature, variants of architectures of our model are constructed to explore the impacts and mutual interactions among the visual representation, textual representations, and the output of the middle LSTM layer. Particularly, the framework is training under a reinforcement learning method to address the exposure bias problem between the training and the testing by the policy gradient optimization. Qualitative analyses indicate the process that our model "translates" image to sentence and further visualization presents the evolution of the hidden states from different hierarchical LSTMs over time. Extensive experiments demonstrate that our model outperforms current state-of-the-art models on three benchmark datasets: Flickr8K, Flickr30K, and MSCOCO. On both image captioning and

关键词： Visualization Semantics Hidden Markov models Decoding Logic gates Training Computer architecture Deep hierarchical structure encoder-decoder LSTM image captioning retrieval vision-sentence

来源：评论

学校读者我要写书评

暂无评论

Deep encoder-decoder Neural Networks for Retinal Blood Vessels Dense Prediction

引用

INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE SYSTEMS 2021年第1期14卷 1078-1086页

作者： Zhang, Wenlu Li, Lusi Cheong, Vincent Fu, Bo Aliasgari, Mehrdad Calif State Univ Long Beach Dept Comp Engn & Comp Sci 1250 Bellflower Blvd Long Beach CA 90840 USA Calif State Univ Los Angels Dept Comp Informat Syst 5151 State Univ Dr Los Angeles CA 90032 USA

Automatic segmentation of retinal blood vessels from fundus images is of great importance in assessing the condition of vascular network in human eyes. The task is primary challenging due to the low contrast of images, the variety of vessels and potential pathology. Previous studies have proposed shallow machine learning based methods to tackle the problem. However, these methods require specific domain knowledge, and the efficiency and robustness of these methods are not satisfactory for medical diagnosis. In recent years, deep learning models have made great progress in various segmentation tasks. In particular, Fully Convolutional Network and U-net have achieved promising results in end-to-end dense prediction tasks. In this study, we propose a novel encoder-decoder architecture based on the vanilla U-net architecture for retinal blood vessels segmentation. The proposed deep learning architecture integrates hybrid dilation convolutions and pixel transposed convolutions in the encoder-decoder model. Such design enables global dense feature extraction and resolves the common "gridding" and "checkerboard" issues in the regular U-net. Furthermore, the proposed network can be efficiently and directly implemented for any semantic segmentation applications. We evaluate the proposed network on two retinal blood vessels data sets. The experimental results show that our proposed model outperforms the baseline U-net model. (C) 2021 The Authors. Published by Atlantis Press B.V.

关键词： Deep learning encoder-decoder Retinal blood vessel Dense prediction

来源：评论

学校读者我要写书评

暂无评论

Code generation from a graphical user interface via attention-based encoder-decoder model

引用

MULTIMEDIA SYSTEMS 2022年第1期28卷 121-130页

作者： Chen, Wen-Yin Podstreleny, Pavol Cheng, Wen-Huang Chen, Yung-Yao Hua, Kai-Lung Natl Taipei Univ Educ Dept Arts & Design Taipei Taiwan Natl Taiwan Univ Sci & Technol Dept Comp Sci & Informat Engn Taipei Taiwan Natl Chiao Tung Univ Inst Elect Hsinchu Taiwan Natl Taiwan Univ Sci & Technol Dept Elect & Comp Engn Taipei Taiwan

Code generation from graphical user interface images is a promising area of research. Recent progress on machine learning methods made it possible to transform user interface into the code using several methods. The encoder-decoder framework represents one of the possible ways to tackle code generation tasks. Our model implements the encoder-decoder framework with an attention mechanism that helps the decoder to focus on a subset of salient image features when needed. Our attention mechanism also helps the decoder to generate token sequences with higher accuracy. Experimental results show that our model outperforms previously proposed models on the pix2code benchmark dataset.

关键词： Code generation encoder-decoder Attention Graphical user interface

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：