检索结果-内蒙古大学图书馆

Deep learning-based concrete defects classification and detection using semantic segmentation

STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL 2024年第1期23卷 383-409页

作者： Arafin, Palisa Billah, A. H. M. Muntasir Issa, Anas Lakehead Univ Dept Civil Engn Thunder Bay ON Canada Univ Calgary Dept Civil Engn Calgary AB Canada United Arab Emirates Univ Dept Civil & Environm Engn Abu Dhabi U Arab Emirates Univ Calgary Dept Civil Engn 2500 Univ Dr NW Calgary AB T2N 1N4 Canada

Visual damage detection of infrastructure using deep learning (DL)-based computational approaches can facilitate a potential solution to reduce subjectivity yet increase the accuracy of the damage diagnoses and accessibility in a structural health monitoring (SHM) system. However, despite remarkable advances with DL-based SHM, the most significant challenges to achieving the real-time implication are the limited available defects image databases and the selection of DL networks depth. To address these challenges, this research has created a diverse dataset with concrete crack (4087) and spalling (1100) images and used it for damage condition assessment by applying convolutional neural network (CNN) algorithms. CNN-classifier models are used to identify different types of defects and semantic segmentation for labeling the defect patterns within an image. Three CNN-based models-Visual Geometry Group (VGG)19, ResNet50, and InceptionV3 are incorporated as CNN-classifiers. For semantic segmentation, two encoder-decoder models, U-Net and pyramid scene parsing network architecture are developed based on four backbone models, including VGG19, ResNet50, InceptionV3, and EfficientNetB3. The CNN-classifier models are analyzed on two optimizers-stochastic gradient descent (SGD), root mean square propagation (RMSprop), and learning rates-0.1, 0.001, and 0.0001. However, the CNN-segmentation models are analyzed for SGD and adaptive moment estimation, trained with three different learning rates-0.1, 0.01, and 0.0001, and evaluated based on accuracy, intersection over union, precision, recall, and F1-score. InceptionV3 achieves the best performance for defects classification with an accuracy of 91.98% using the RMSprop optimizer. For crack segmentation, EfficientNetB3-based U-Net, and for spalling segmentation, IncenptionV3-based U-Net model outperformed all other algorithms, with an F1-score of 95.66 and 89.43%, respectively.

关键词： Concrete defects convolutional neural network semantic segmentation encoder-decoder model structural health monitoring

来源：评论

学校读者我要写书评

暂无评论

Boosting Just-In-Time Code Comment Updating Via Programming Context and Refactor

引用

INTERNATIONAL JOURNAL OF SOFTWARE ENGINEERING AND KNOWLEDGE ENGINEERING 2023年第10期33卷 1619-1649页

作者： Mi, Xiangbo Zhang, Jingxuan Tang, Yixuan Ju, Yue Lan, Jinpeng Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing Peoples R China

Comments are summary descriptions of code snippets. When analyzing and maintaining programs, developers tend to read tidy comments rather than lengthy code. To prevent developers from misunderstanding the program or leading to potential bugs, ensuring the consistency and co-evolution of comments and the corresponding code is an integral development activity in practice. Nevertheless, when modifying code, developers sometimes neglect to update the relevant comments, resulting in inconsistency. Such comments may pose threats to the comprehension and maintenance of the software. In our study, we propose an overall approach named Context and Refactor based Comment Updater (CRCU). CRCU is a Just-In-Time (JIT) comment updater for specific commits. It takes a commit-id as input and updates all the method comments in this commit according to the code change. CRCU could be viewed as an optimization and augmentation of existing comment updaters, especially those that rely only on neural networks. Compared to the existing comment updaters, CRCU fully leverages the programming context and refactoring types of the modified methods to improve its performance. In addition, several customized enhancements in data pre-processing are introduced in CRCU to handle and filter out low-quality commits. We conduct extensive experiments to evaluate the effectiveness of CRCU. The evaluation results show that CRCU combined with the state-of-the-art approaches could improve the average Accuracy by 6.87% and reduce the developers' edits by 0.298 on average.

关键词： Code-comment co-evolution refactoring type detection programming context detection encoder-decoder model

来源：评论

学校读者我要写书评

暂无评论

Tulu Language Text Recognition and Translation

引用

IEEE ACCESS 2024年 12卷 12734-12744页

作者： Prathwini Rodrigues, Anisha P. Vijaya, P. Fernandes, Roshan NITTE NMAM Inst Technol Dept Master Comp Applicat Nitte 574110 India NITTE NMAM Inst Technol Dept Comp Sci & Engn Nitte 574110 India Modern Coll Business & Sci Dept Math & Comp Sci Muscat 133 Oman

Language is a primary means of communication, but it is not the only means;knowing a language does, however, assist speed up the process. Many distinct languages are spoken worldwide, and people use them to communicate. This is only one of the many reasons why language is so crucial. Based on the literature survey, it is evident that there is a lack of available translators for the Tulu language. Despite being prevalent predominantly in Karnataka, the Tulu language has not been as widely spoken as other Indian languages until recently, although it gained enough recognition to become the second language in Karnataka. The purpose of our research work aims at translating the English language into the Tulu language. During the evaluation the system was tested on a dataset consisting of handwritten characters during the evaluation process Convolutional Neural Networks used achieved an accuracy rate of 92%. To translate English to the Tulu language, we employed a parallel sentence dataset for the neural approach and a parallel word dataset for the rule-based approach. The rule-based approach resulted in an 89% accuracy rate for word-based analysis and an 81% accuracy rate for sentence-based analysis for the English-to-Tulu language translation. The neural machine translation approach of the encoder-decoder model with LSTM is been used to accomplish translation from English to Tulu with a BLEU score of 0.83 and Tulu to English with a BLUE score of 0.65. The model also employed hybrid machine translation to enhance the translation.

关键词： Character recognition Machine translation Principal component analysis Machine learning algorithms Surveys Linguistics Handwriting recognition Neural networks Convolutional neural networks Long short term memory rule-based method neural network translation convolutional neural network (CNN) encoder-decoder model long short-term memory (LSTM)

来源：评论

学校读者我要写书评

暂无评论

DAFA: Dialog System Domain Adaptation With a Filter and an Amplifier

引用

IEEE ACCESS 2020年 8卷 45041-45049页

作者： Yu, Jianfeng Yang, Yan Chen, Chengcai He, Liang Yu, Zhou China Normal Univ Dept Comp Sci Shanghai 200062 Peoples R China Xiaoi Res Shanghai 201803 Peoples R China Univ Calif Davis Dept Comp Sci Davis CA 95616 USA

End-to-end task-oriented dialog systems have attracted vast amounts of attention in recent years, mainly because of their ease of training. However, such an end-to-end model requires a large number of labeled dialogs to train. Labeled dialogs are always difficult to obtain in real-world settings. We propose a domain adaptive end-to-end task-oriented dialog model that transfers knowledge in source domains to a target domain with limited training samples. Specifically, we design a domain adaptive filter in the encoder-decoder model to reduce useless features in source domains and preserve common features. A domain adaptive amplifier is designed to enhance the target domain impact. We evaluate our method on both synthetic dialog and human-human dialog datasets and achieve state-of-the-art results.

关键词： Adaptation models Task analysis Training Decoding Training data Data models Computer science Neural network dialogue system domain adaptation encoder-decoder model

来源：评论

学校读者我要写书评

暂无评论

Joint Object Affordance Reasoning and Segmentation in RGB-D Videos

引用

IEEE ACCESS 2021年 9卷 89699-89713页

作者： Thermos, Spyridon Potamianos, Gerasimos Daras, Petros Univ Edinburgh Sch Engn Edinburgh EH9 3JL Midlothian Scotland Univ Thessaly Dept Elect & Comp Engn Volos 38221 Greece Informat Technol Inst Ctr Res & Technol Hellas Visual Comp Lab Thessaloniki 57001 Greece

Understanding human-object interaction is a fundamental challenge in computer vision and robotics. Crucial to it is the ability to infer "object affordances" from visual data, namely the types of interaction supported by an object of interest and the object parts involved. Such inference can be approached as an "affordance reasoning" task, where object affordances are recognized and localized as image heatmaps, and as an "affordance segmentation" task, where affordance labels are obtained at a more detailed, image pixel level. To tackle the two tasks, existing methods typically: (i) treat them independently;(ii) adopt static image-based models, ignoring the temporal aspect of human-object interaction;and / or (iii) require additional strong supervision concerning object class and location. In this paper, we focus on both tasks, while addressing all three aforementioned shortcomings. For this purpose, we propose a deep-learning based dual encoder-decoder model for joint affordance reasoning and segmentation, which learns from our recently introduced SOR3D-AFF corpus of RGB-D human-object interaction videos, without relying on object localization and classification. The basic components of the model comprise: (i) two parallel encoders that capture spatio-temporal interaction information;(ii) a reasoning decoder that predicts affordance heatmaps, assisted by an affordance classifier and an attention mechanism;and (iii) a segmentation decoder that exploits the predicted heatmap to yield pixel-level affordance segmentation. All modules are jointly trained, while the system can operate on both static images and videos. The approach is evaluated on four datasets, surpassing the current state-of-the-art in both affordance reasoning and segmentation.

关键词： Affordances Cognition Decoding Task analysis Image segmentation Heating systems Videos Object affordances human-object interaction reasoning semantic segmentation deep learning encoder-decoder model attention mechanism RGB-D video

来源：评论

学校读者我要写书评

暂无评论

Improving the readability and saliency of abstractive text summarization using combination of deep neural networks equipped with auxiliary attention mechanism

引用

JOURNAL OF SUPERCOMPUTING 2022年第2期78卷 2528-2555页

作者： Aliakbarpour, Hassan Manzuri, Mohammad Taghi Rahmani, Amir Masoud Islamic Azad Univ Dept Comp Engn Sci & Res Branch Tehran Iran Sharif Univ Technol Comp Engn Dept Tehran Iran

Rapid and exponential development of textual data in recent years has yielded to the need for automatic text summarization models which aim to automatically condense a piece of text into a shorter version. Although various unsupervised and machine learning-based approaches have been introduced for text summarization during the last decades, the emergence of deep learning has made remarkable progress in this field. However, deep learning-based text summarization models are still in their early steps of development and their potential has yet to be fully explored. Accordingly, a novel abstractive summarization model is proposed in this paper which utilized the combination of convolutional neural network and long short-term memory integrated with auxiliary attention in its encoder to increase the saliency and coherency of generated summaries. The proposed model was validated on CNN\Daily Mail and DUC-2004 datasets and empirical results indicated that not only the proposed model outperformed existing models in terms of ROUGE metric but also its generated summaries had higher saliency and readability compared to the baseline model according to human evaluation.

关键词： Abstractive text summarization encoder-decoder model Convolutional neural network Long short-term memory Auxiliary attention

来源：评论

学校读者我要写书评

暂无评论

CNN-based encoder-decoder networks for salient object detection: A comprehensive review and recent advances

引用

INFORMATION SCIENCES 2021年 546卷 835-857页

作者： Ji, Yuzhu Zhang, Haijun Zhang, Zhao Liu, Ming Harbin Inst Technol Dept Comp Sci Shenzhen Peoples R China Hefei Univ Technol Dept Comp Sci Hefei Peoples R China Harbin Inst Technol Sch Astronaut Harbin Peoples R China

Convolutional neural network (CNN)-based encoder-decoder models have profoundly inspired recent works in the field of salient object detection (SOD). With the rapid development of encoder-decoder models with respect to most pixel-level dense prediction tasks, an empirical study still does not exist that evaluates performance by applying a large body of encoder-decoder models on SOD tasks. In this paper, instead of limiting our survey to SOD methods, a broader view is further presented from the perspective of fundamental architectures of key modules and structures in CNN-based encoder-decoder models for pixel-level dense prediction tasks. Moreover, we focus on performing SOD by leveraging deep encoder-decoder models, and present an extensive empirical study on baseline encoder-decoder models in terms of different encoder backbones, loss functions, training batch sizes, and attention structures. Moreover, state-of-the-art encoder-decoder models adopted from semantic segmentation and deep CNN-based SOD models are also investigated. New baseline models that can outperform state-of-the-art performance were discovered. In addition, these newly discovered baseline models were further evaluated on three video-based SOD benchmark datasets. Experimental results demonstrate the effectiveness of these baseline models on both imageand video-based SOD tasks. This empirical study is concluded by a comprehensive summary which provides suggestions on future perspectives. (c) 2020 Elsevier Inc. All rights reserved.

关键词： Salient object detection encoder-decoder model Pixel-level classification Video saliency Empirical study

来源：评论

学校读者我要写书评

暂无评论

Gas concentration prediction based on ED-SLSTM model under the framework of Trend Prediction-Time Point Prediction

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2024年第10期15卷 4695-4707页

作者： Wang, Xiangqian Xu, Ningke Meng, Xiangrui Anhui Univ Sci & Technol Sch Econ & Management Huainan 232000 Peoples R China Anhui Univ Sci & Technol Sch Comp Sci & Engn Huainan 232000 Peoples R China

In order to solve the problems of low prediction accuracy and inability to achieve long-term prediction in traditional coal mine gas concentration prediction, this paper proposes a new gas concentration prediction framework based on the problem of independent gate units of traditional Long Short Time Memory (LSTM) neural network-"Trend Prediction-Time Point Prediction" (TP-TPP) structure, and it is the first time to add a selection gate mechanism to the LSTM model to construct the Select LSTM (SLSTM) model, which is in line with the actual production pattern of coal mines. Firstly, the data of gas multi-parameter time series are processed, and then the input data features under the framework of trend prediction are reconstructed. Then the modified Snake Optimizer (SO) algorithm is used to simplify the parameter adjustment process of neural network. Here the encoder-decoder-SLSTM (ED-SLSTM) model is compared with LightGBM, LSTM, Bi-LSTM, LSTM-Attention and CEEMDAN-LSTM models under the framework of TP-TPP, respectively. The performance of each evaluation index in ED-SLSTM is closer to the real value than them of other models. The results show that ED-SLSTM has higher prediction accuracy and better prediction effect.

关键词： Gas concentration prediction TP-TPP SLSTM encoder-decoder model Snake optimizer

来源：评论

学校读者我要写书评

暂无评论

A Multilevel Transfer Learning Technique and LSTM Framework for Generating Medical Captions for Limited CT and DBT Images

引用

JOURNAL OF DIGITAL IMAGING 2022年第3期35卷 564-580页

作者： Aswiga, R., V Shanthi, A. P. Amrita Vishwa Vidyapeetham Amrita Sch Engn Dept Comp Sci & Engn Chennai 601103 Tamil Nadu India Anna Univ Coll Engn Dept Comp Sci & Engn Guindy CEG Chennai 600025 Tamil Nadu India

Medical image captioning has been recently attracting the attention of the medical community. Also, generating captions for images involving multiple organs is an even more challenging task. Therefore, any attempt toward such medical image captioning becomes the need of the hour. In recent years, the rapid developments in deep learning approaches have made them an effective option for the analysis of medical images and automatic report generation. But analyzing medical images that are scarce and limited is hard, and it is difficult even with machine learning approaches. The concept of transfer learning can be employed in such applications that suffer from insufficient training data. This paper presents an approach to develop a medical image captioning model based on a deep recurrent architecture that combines Multi Level Transfer Learning (MLTL) framework with a Long Short-Term-Memory (LSTM) model. A basic MLTL framework with three models is designed to detect and classify very limited datasets, using the knowledge acquired from easily available datasets. The first model for the source domain uses the abundantly available non-medical images and learns the generalized features. The acquired knowledge is then transferred to the second model for the intermediate and auxiliary domain, which is related to the target domain. This information is then used for the final target domain, which consists of medical datasets that are very limited in nature. Therefore, the knowledge learned from a non-medical source domain is transferred to improve the learning in the target domain that deals with medical images. Then, a novel LSTM model, which is used for sequence generation and machine translation, is proposed to generate captions for the given medical image from the MLTL framework. To improve the captioning of the target sentence further, an enhanced multi-input Convolutional Neural Network (CNN) model along with feature extraction techniques is proposed. This enhanced multi-in

关键词： Medical image caption Deep learning encoder-decoder model Transfer learning

来源：评论

学校读者我要写书评

暂无评论

AENet: attention enhancement network for industrial defect detection in complex and sensitive scenarios

引用

JOURNAL OF SUPERCOMPUTING 2024年第9期80卷 11845-11868页

作者： Wan, Yi Yi, Lingjie Jiang, Bo Chen, Junfan Jiang, Yi Xie, Xianzhong Chongqing Univ Posts & Telecommun Coll Comp Sci & Technol Chongqing 400065 Peoples R China

Conventional image processing and machine learning based on handcrafted features struggle to meet the real time and high-accuracy requirements for industrial defect detection in complex, sensitive, and dynamic environments. To address this issue, this paper proposes AENet, a novel real-time defect detection network based on an encoder-decoder model, which achieves high detection accuracy and efficiency while demonstrating excellent convergence and generalization. Firstly, a spatial channel attention module in the encoding network is designed to exploit both spatial attention and channel attention using a multi-head 3D self-attention mechanism. This improves parallelism and detection efficiency. Secondly, the decoding network of AENet incorporates the cross-level attention fusion module, which fuses input features from different layers. Combined with multi-level upsampling design, the decoder enhances the representation of defect details. Furthermore, we insert a simplified aggregator into the encoder-decoder network to extract feature information at different scales with low computational cost. This aggregation process aids in training and inference on industrial defect datasets by incorporating contextual information. Extensive experimental results demonstrate that AENet outperforms other segmentation models in accomplishing defect recognition and segmentation in challenging optical environments. It exhibits a faster convergence than other networks and a balance between accuracy and speed. It achieves a recognition accuracy of over 96% for almost all types of defects in the actual industrial environment on the NVIDIA Tesla V100 GPU.

关键词： encoder-decoder model Attention mechanism Attention fusion Industrial defect detection

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：