检索结果-内蒙古大学图书馆

Mathematical expression recognition using a new deep neural model

NEURAL NETWORKS 2023年 167卷 865-874页

作者： Mirkazemy, Abolfazl Adibi, Peyman Ehsani, Seyed Mohhamad Saied Darvishy, Alireza Hutter, Hans-Peter Univ Isfahan Fac Comp Engn Artificial Intelligence Dept Esfahan Iran Zurich Univ Appl Sci ZHAW Sch Engn Winterthur Switzerland

In this paper, we propose a novel deep neural model for Mathematical Expression Recognition (MER). The proposed model uses encoder-decoder transformer architecture that is supported by additional pre/post-processing modules, to recognize the image of mathematical formula and convert it to a well-formed language. A novel pre-processing module based on domain prior knowledge is proposed to generate random pads around the formula's image to create more efficient feature maps and keeps all the encoder neurons active during the training process. Also, a new post-processing module is developed which uses a sliding window to extract additional position-based information from the feature map, that is proved to be useful in the recognition process. The recurrent decoder module uses the combination of feature maps and the additional position-based information, which takes advantage of a soft attention mechanism, to extract the formula context into the LaTeX well-formed language. Finally, a novel Reinforcement Learning (RL) module processes the decoder output and tunes its results by sending proper feedbacks to the previous steps. The experimental results on im2latex100k benchmark dataset indicate that each devised pre/post-processing as well as the RL refinement module has a positive effect on the performance of the proposed model. The results also demonstrate the higher accuracy of the proposed model compared to the state-of-the-art methods. (c) 2023 Elsevier Ltd. All rights reserved.

关键词： Mathematical expression recognition Deep learning encoder-decoder architecture Attention Scientific documents accessibility

来源：评论

学校读者我要写书评

暂无评论

Adaptive Bus Encoding for Transition Reduction on Off-Chip Buses With Dynamically Varying Switching Characteristics

引用

IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS 2017年第11期25卷 3057-3066页

作者： Sarkar, Sumantra Biswas, Ayan Dhar, Anindya Sundar Rao, Rahul M. IIT Kharagpur Dept Elect & Elect Commun Engn Kharagpur 721302 W Bengal India IBM India Private Ltd Bangalore 560045 Karnataka India

This paper presents an adaptive encoding framework for the reduction of transition activity in high-capacitance off-chip data buses, since power dissipation associated with those buses can be significant for high-speed communication. The technique relies on the observation of data characteristics over fixed window sizes and formation of cluster with bit lines having highly correlated switching patterns. The proposed method utilizes redundancy in space and time to prevent loss of information while retrieving data. We present analytical and experimental analyses, which demonstrate the activity reduction of our encoding scheme for various data. The extra power cost due to the encoder and decoder circuitry along with redundancy is offset due to reduced number of off-chip transitions.

关键词： Bus encoding schemes encoder-decoder architecture low power VLSI design off-chip bus

来源：评论

学校读者我要写书评

暂无评论

Pavement crack detection network based on pyramid structure and attention mechanism

引用

IET IMAGE PROCESSING 2020年第8期14卷 1580-1586页

作者： Xiang, Xuezhi Zhang, Yuqi El Saddik, Abdulmotaleb Harbin Engn Univ Sch Informat & Commun Engn Harbin 150001 Peoples R China Univ Ottawa Sch Elect Engn & Comp Sci Ottawa ON K1N 6N5 Canada

Automatic detection of pavement crack is an important task for conducting road maintenance. However, as an important part of the intelligent transportation system, automatic pavement crack detection is challenging due to the poor continuity of cracks, the different width of cracks, and the low contrast between cracks and the surrounding pavement. This study proposes a novel pavement crack detection method based on an end-to-end trainable deep convolution neural network. The authors build the network using the encoder-decoder architecture and adopt a pyramid module to exploit global context information for the complex topology structures of cracks. Moreover, they introduce a spatial-channel combinational attention module into the encoder-decoder network for refining crack features. Further, the dilated convolution is used to reduce the loss of crack details due to the pooling operation in the encoder network. In addition, they introduce a lovasz hinge loss function, which is suitable for small objects. They train the authors' network on the CRACK500 dataset and evaluate it on three pavement crack datasets. Among the methods they compare, their method can achieve the best experimental results.

关键词： learning (artificial intelligence) object detection crack detection roads structural engineering computing image recognition cracks maintenance engineering convolutional neural nets intelligent transportation systems pavement crack detection network pyramid structure attention mechanism automatic pavement crack detection surrounding pavement spatial-channel combinational attention module crack features CRACK500 dataset pavement crack datasets encoder-decoder network road maintenance encoder-decoder architecture end-to-end trainable deep convolution neural network complex topology structures pooling operation intelligent transportation system

来源：评论

学校读者我要写书评

暂无评论

Seismic Facies Analysis Based on Deep Learning

引用

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 2020年第7期17卷 1119-1123页

作者： Zhang, Yuxi Liu, Yang Zhang, Haoran Xue, Hao China Univ Petr State Key Lab Petr Resources & Prospecting Beijing 102249 Peoples R China China Univ Petr Sch Petr Karamay 834000 Peoples R China CNOOC Int Ltd Beijing 100028 Peoples R China

Seismic facies analysis is to study the sedimentary environment of stratigraphic sequence and provides an important basis for reservoir prediction. Most of the existing analysis methods have low efficiency and heavily rely on manual experience, and therefore, it is difficult to interpret increasingly complex seismic data. Deep learning techniques can help to solve these problems and achieve automatic seismic facies classification. We regard seismic facies classification as a target segmentation problem and propose new method and training strategies. Our workflow primarily involves four sections. First, we process the manually annotated labels and seismic data with mirroring and cropping operations to ensure that network can accept input with arbitrary size and the model training is not limited to GPU memory. Second, data augmentation is applied to automatically generate massive training samples from the processed data. Third, we build two independent networks based on encoder-decoder architecture: one identifies all seismic facies simultaneously, and the other identifies single seismic facies in each model. However, both the results of the two networks have some drawbacks. Fourth, to overcome these drawbacks, we propose an ensemble learning method to get optimized model and test it on 3-D seismic data. The testing results manifest that the proposed method can improve the predictive ability of model, accurately describe the seismic facies, and can be applicable to entire seismic data volume.

关键词： Training Testing Data models Deep learning Reservoirs Graphics processing units Decoding Deep learning encoder-decoder architecture ensemble learning seismic facies analysis

来源：评论

学校读者我要写书评

暂无评论

Deep Road Scene Understanding

引用

IEEE SIGNAL PROCESSING LETTERS 2019年第4期26卷 587-591页

作者： Zhou, Wujie Lv, Sijia Jiang, Qiuping Yu, Lu Zhejiang Univ Sci & Technol Sch Informat & Elect Engn Hangzhou 310023 Zhejiang Peoples R China Zhejiang Univ Inst Informat & Commun Engn Hangzhou 310027 Zhejiang Peoples R China Ningbo Univ Fac Informat Sci & Engn Ningbo 315211 Zhejiang Peoples R China

Road scene understanding is a difficult task in autonomous driving. In this letter, we propose a novel deep encoder-decoder architecture for road scene understanding in an end-to-end manner. This core trainable understanding engine includes an encoder network, a decoder network with two streams, and a pixel-level fusion network with classification layer. The encoder network is composed of the front-end model of the classical convolution neural network, VGGNet. The decoder network with two streams includes multi-scale skip connection modules to reduce the down-scaling effect. Finally, a fusion network fuses the two-level information from the two streams of the decoder network for precise pixel-level classification. Additionally, the convolution layer is added to each skip connection module to increase the depth of the architecture. Our architecture achieves outstanding performance on the publicly available Cam Vid dataset and significantly outperforms previous architectures. This deep architecture is ideal for road scene understanding.

关键词： Road scene understanding encoder-decoder architecture skip connection fusion layer

来源：评论

学校读者我要写书评

暂无评论

Data Repair Without Prior Knowledge Using Deep Convolutional Neural Networks

引用

IEEE ACCESS 2020年 8卷 105351-105361页

作者： Qie, Youtian Song, Ping Hao, Chuangbo Beijing Inst Technol Key Lab Biomimet Robots & Syst Minist Educ Beijing 100081 Peoples R China Beijing Jinghang Computat & Commun Res Inst Beijing 100074 Peoples R China

In recent years, the use of wireless sensor networks has become increasingly widespread. Because of the instability of wireless networks, packet loss occasionally occurs. To reduce the impact of packet loss on data integrity, we take advantage of the deep neural network's excellent ability to understand natural data and propose a data repair method based on a deep convolutional neural network with an encoder-decoder architecture. Compared with common interpolation algorithms and compressed sensing algorithms, this method obtains better repair results, is suitable for a wider range of applications, and does not need prior knowledge. This method adopts measures such as preparing training set data as well as the design and optimization of loss functions to achieve faster convergence speed, higher repair accuracy, and better stability. To fairly compare the repair performance of different signals, the mean squared error, relative peak-to-peak average error, and relative peak-to-peak max error are adopted to quantitatively evaluate the repair results of different signals. Comparative experiments prove that this method has better data recovery performance than traditional interpolation and compressed sensing algorithms.

关键词： Data repair method without prior knowledge encoder-decoder architecture deep convolutional neural network wireless sensor networks evaluation method

来源：评论

学校读者我要写书评

暂无评论

ACEANet: Ambiguous Context Enhanced Attention Network for skin lesion segmentation

引用

INTELLIGENT DATA ANALYSIS 2024年第3期28卷 791-805页

作者： Jiang, Yun Qiao, Hao Northwest Normal Univ Dept Comp Sci & Engn Lanzhou Gansu Peoples R China

Skin lesion segmentation from dermatoscopic images is essential for the diagnosis of skin cancer. However, it is still a challenging task due to the ambiguity of the skin lesions, the irregular shape of the lesions and the presence of various interfering factors. In this paper, we propose a novel Ambiguous Context Enhanced Attention Network (ACEANet) based on the classical encoder-decoder architecture, which is able to accurately and reliably segment a variety of lesions with efficiency. Specifically, a novel Ambiguous Context Enhanced Attention module is embedded in the skip connection to augment the ambiguous boundary information. A Dilated Gated Fusion block is employed in the end of the encoding phase, which effectively reduces the loss of spatial location information due to continuous downsampling. In addition, we propose a novel Cascading Global Context Attention to fuse feature information generated by the encoder with features generated by the decoder of the corresponding layer. In order to verify the effectiveness and advantages of the proposed network, we have performed comparative experiments on ISIC2018 dataset and PH2 dataset. Experiments results demonstrate that the proposed model has superior segmentation performance for skin lesions.

关键词： Skin lesion segmentation medical image processing feature extraction encoder-decoder architecture

来源：评论

学校读者我要写书评

暂无评论

Semantic segmentation of satellite images for crop type identification in smallholder farms

引用

JOURNAL OF SUPERCOMPUTING 2024年第2期80卷 1367-1395页

作者： Buttar, Preetpal Kaur Sachan, Manoj Kumar St Longowal Inst Engn & Technol Dept Comp Sci & Engn Longowal 148106 Punjab India

Accurate and reliable crop type identification from satellite images provides a foundation for crop yield predictions which paves the way to help ensure food security. Most of the work done in the field of crop type mapping using remote sensing is restricted to the developed countries having large field parcels, while a little effort has been directed towards doing so for developing countries, where this task becomes more challenging due to the small size of field parcels, irregular shapes of the fields, and an acute shortage of labelled datasets for training supervised machine learning models. In this research, we try to fill this gap in the literature by exploring the feasibility of performing the semantic segmentation of agricultural fields from satellite images by proposing an encoder-decoder-based semantic segmentation architecture, CropNet, with a ResNet network as the encoder backbone and the use of attention modules in the decoder to allow the model to focus on more important portions of the feature maps and the feature fusion to concatenate the feature maps from all the decoder nodes getting a more precise prediction by bringing the spatial location information from the previous layers. The architecture outperformed the state of the art by 0.51% and 1.3%, on overall accuracy and macro-F1 score, respectively, after being trained on the "2019 Zindi's Farm Pin Crop Detection" dataset of Sentinel-2 images. The model achieved a field-wise overall classification accuracy of 78.06% and macro-F1 score of 67.3% and a pixel-wise segmentation mean Intersection over Union (mIoU) of 62.22% which is an improvement of 2.56% over the state-of-the-art methods, thereby demonstrating that our model is computationally efficient for the job of semantic segmentation of crop types from the satellite images in the difficult scenario of smallholder farms.

关键词： Semantic segmentation Remote sensing encoder-decoder architecture Sentinel-2 Crop type mapping Smallholder farms

来源：评论

学校读者我要写书评

暂无评论

Video Salient Object Detection Via Multi-level Spatiotemporal Bidirectional Network Using Multi-scale Transfer Learning

引用

IETE JOURNAL OF RESEARCH 2024年第11期70卷 8077-8088页

作者： Sharma, Gaurav Singh, Maheep Berwal, Krishan NIT Uttarakhand Srinagar Garhwal Dept Comp Sci Engn Srinagar 246174 Uttarakhand India Mil Coll Telecommun Engn Mhow Madhya Pradesh India

Video saliency prediction aims to resemble human visual attention by identifying the most relevant and significant elements in a video frame or sequence. This task becomes notably intricate in scenarios characterized by dynamic elements such as rapid motion, occlusions, blur, background variations, and nonrigid deformations. Therefore, the inherent complexity of human visual attention behavior during dynamic scenes necessitates the assessment of both temporal and spatial data. Existing video saliency frameworks often falter under such conditions, and relying solely on image saliency models neglects crucial temporal information in videos. This study presents a new Video Salient Object Detection via Multi-level Spatiotemporal Bidirectional Network using Multi-scale Transfer Learning (MSB-Net) to address the problem of identifying significant objects in videos. The proposed MSB-Net achieves notable results for a given sequence of frames by employing multi-scale transfer learning with an encoder and decoder approach to acquire knowledge and saliency map attributes spatially and temporally. The proposed MSB-Net model has bidirectional LSTM (Long Short-Term Memory) and CNN (Convolutional Neural Network) components. The VGG16 (Video Geometry Group) and VGG19 architectures extract multi-scale features from the input video frames. Evaluation of diverse datasets, namely DAVIS-T, SegTrack-V2, ViSal, VOS-T, and DAVSOD-T, demonstrates the model's effectiveness, outperforming other competitive models based on parameters such as MAE, F-measure, and S-measure.

关键词： encoder-decoder architecture Saliency Spatiotemporal Video

来源：评论

学校读者我要写书评

暂无评论

Lightweight and efficient asymmetric network design for real-time semantic segmentation

引用

APPLIED INTELLIGENCE 2022年第1期52卷 564-579页

作者： Zhang, Xiu-Ling Du, Bing-Ce Luo, Zhao-Ci Ma, Kai Yanshan Univ Qinhuangdao Minist Educ Intelligent Control Syst & Intelligen Engn Res Ctr Qinhuangdao Hebei Peoples R China Yanshan Univ Qinhuangdao Key Lab Ind Comp Control Engn Hebei Prov Qinhuangdao Hebei Peoples R China

With the increasing demand for application scenarios such as autonomous driving and drone aerial photography, it has become a challenging problem that how to achieve the best trade-off between segmentation accuracy and inference speed while reducing the number of parameters. In this paper, a lightweight and efficient asymmetric network (LEANet) for real-time semantic segmentation is proposed to address this problem. Specifically, LEANet adopts an asymmetric encoder-decoder architecture. In the encoder, a depth-wise asymmetric bottleneck module with separation and shuffling operations (SS-DAB module) is proposed to jointly extract local and context information. In the decoder, a pyramid pooling module based on channel-wise attention (CA-PP module) is proposed to aggregate multi-scale context information and guide feature selection. Without any pre-training and post-processing, LEANet respectively achieves the accuracy of 71.9% and 67.5% mean Intersection over Union (mIoU) with the speed of 77.3 and 98.6 Frames Per Second (FPS) on the Cityscapes and CamVid test sets. These experimental results show that LEANet achieves an optimal trade-off between segmentation accuracy and inference speed with only 0.74 million parameters.

关键词： Real-time Semantic segmentation Lightweight network Asymmetric network encoder-decoder architecture

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：