检索结果-内蒙古大学图书馆

Enhancing eyeglasses removal in facial images: a novel approach using translation models for eyeglasses mask completion

引用

Multimedia Tools and Applications 2024年 1-20页

作者： Esmaily, Zahra Ebrahimpour-Komleh, Hossein Department of Electrical and Computer Engineering University of Kashan Kashan Iran

Accurately removing eyeglasses from facial images is crucial for improving the performance of various face-related tasks such as verification, identification, and reconstruction. This paper presents a novel approach to enhancing eyeglasses removal by integrating a mask completion technique into the existing framework. Our method focuses on improving the accuracy of eyeglasses masks, which is essential for subsequent eyeglasses and shadow removal steps. We introduce a unique dataset specifically designed for eyeglasses mask image completion. This dataset is generated by applying Top-Hat morphological operations to existing eyeglasses mask datasets, creating a collection of images containing eyeglasses masks in two states: damaged (incomplete) and complete (ground truth). A Pix2Pix image-to-image translation model is trained on this newly created dataset for the purpose of restoring incomplete eyeglass mask predictions. This restoration step significantly improves the accuracy of eyeglass frame extraction and leads to more realistic results in subsequent eyeglasses and shadow removal. Our method incorporates a post-processing step to refine the completed mask, preventing the formation of artifacts in the background or outside of the eyeglasses frame box, further enhancing the overall quality of the processed image. Experimental results on CelebA, FFHQ, and MeGlass datasets showcase the effectiveness of our method, outperforming state-of-the-art approaches in quantitative metrics (FID, KID, MOS) and qualitative evaluations. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Motion Planning and Control for Improved Ride Comfort in Urban Autonomous Driving

引用

Journal of Institute of Control, Robotics and Systems 2024年第3期30卷 206-213页

作者： Jang, Kyeongjun Kim, Hakil Department of Electrical and Computer Engineering Inha University Korea Republic of

Self-driving shuttle services frequently provide a less comfortable experience for passengers than human drivers. This observation emphasizes the importance of motion planning and control methods. This paper proposes a method to improve ride comfort in urban autonomous driving. The path information is utilized for velocity planning to minimize lateral acceleration and limiting the acceleration when controlling the PID to avoid rapid deceleration. Additionally, the hybrid lateral controller increases path following performance. The proposed method proved excellent performance compared to the existing geometric lateral control method. A decision tree was employed to evaluate ride comfort to meet ISO standards. The acceleration and jerk were assessed to determine their impact on ride comfort. Furthermore, the lateral error and the heading error were analyzed to evaluate the stability of the route during autonomous driving. An experiment was conducted and verified in the actual urban environment of Songdo New Town, Incheon, to ascertain the efficacy of the proposed method in enhancing ride comfort. © ICROS 2024.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

A Study on Reliability of Interruption Time on Thyristor-based Multi-module SSCB by Parameter Optimization

引用

Transactions of the Korean Institute of electrical Engineers 2024年第1期73卷 37-43页

作者： Kim, Jong-Myoung Song, Jin-Sol Kim, Chul-Hwan Department of Electrical and Computer Engineering Sungkyunkwan University Korea Republic of

With the increasing penetration of PV and RES, the DC source in the power system is rapidly increasing. As a result, there is active discussion regarding the introduction of MVDC distribution networks in Korea. Because DC fault have different characteristics compared to AC fault, research on DC circuit breakers is essential for establishing protection system in MVDC distribution networks. In this paper, research on the optimal parameter design of a SCR-based solid-state circuit breaker is conducted. Proposed SSCB combines the advantages of SSCB with fast interruption time and interrupting method of HCB that generate artificial zero-current crossing point. The proposed SSCB is capable of performing the interruption within 50us upon receiving a trip signal from the relay. It is expected to contribute to the establishment of a reliable protection coordination system in MVDC network. The validation of the operation of the SSCB was performed by Matlab/Simulink. © 2024 Korean Institute of electrical Engineers. All rights reserved.

关键词： Electric power distribution

来源：评论

学校读者我要写书评

暂无评论

Knowledge distillation on spatial-temporal graph convolutional network for traffic prediction

引用

International Journal of computers and Applications 2025年第1期47卷 45-56页

作者： Izadi, Mohammad Safayani, Mehran Mirzaei, Abdolreza Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan Iran

Efficient real-time traffic prediction is crucial for reducing transportation time. To predict traffic conditions, we employ a spatio-temporal graph neural network (ST-GNN) to model our real-time traffic data as temporal graphs. Despite its capabilities, it often encounters challenges in delivering efficient real-time predictions for real-world traffic data. Recognizing the significance of timely prediction due to the dynamic nature of real-time data, we employ knowledge distillation (KD) as a solution to enhance the execution time of ST-GNNs for traffic prediction. In this paper, We introduce a cost function designed to train a network with fewer parameters (the student) using distilled data from a complex network (the teacher) while maintaining its accuracy close to that of the teacher. We use knowledge distillation, incorporating spatial-temporal correlations from the teacher network to enable the student to learn the complex patterns perceived by the teacher. However, a challenge arises in determining the student network architecture rather than considering it inadvertently. To address this challenge, we propose an algorithm that utilizes the cost function to calculate pruning scores, addressing small network architecture search issues, and jointly fine-tunes the network resulting from each pruning stage using KD. Ultimately, we evaluate our proposed ideas on two real-world datasets, PeMSD7 and PeMSD8. The results indicate that our method can maintain the student's accuracy close to that of the teacher, even with the retention of only (Formula presented.) of network parameters. © 2024 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Cost functions

来源：评论

学校读者我要写书评

暂无评论

Predicting Distant Metastatic Sites in Cancer Using miRNA and mRNA Expression Data

引用

Current Bioinformatics 2025年第5期20卷 473-485页

作者： Mamatkarimov, Dostonjon Kang, Jiahui Han, Kyungsook Department of Electrical and Computer Engineering Inha University Incheon Korea Republic of

Background: Cancer patients with metastasis face a much lower survival rate and a higher risk of recurrence than those without metastasis. So far, several learning methods have been proposed to predict cancer metastasis, but most of these methods are intended to predict lymph node metastasis rather than distant metastasis. Distant metastasis is more difficult to predict than lymph node metastasis because distant metastasis is detected after a comprehensive examination of the entire body, and there are not enough publicly available tumor samples with distant metastasis that can be used for training learning methods. Predicting distant metastatic sites is even more challenging than predicting whether distant metastasis will occur or not. Methods: The problem of predicting distant metastatic sites is a multi‐class and multi‐label classification problem;there are more than two classes for distant metastatic sites (bone, brain, liver, lung, and other organs), and a single sample can have multiple labels for multiple metastatic sites. We transformed the multi‐label and multi‐class problem into multiple single‐label binary problems. For each metastatic site, we built a random forest model that deals with binary classification and linked the models along a chain. Results: Testing the model on miRNA and mRNA expression datasets of several cancer types showed a high performance in all performance measures. In the comparison of our model with other methods, our method outperformed the others. Conclusion: We developed a new method for predicting multiple metastatic sites using miRNA and mRNA expression data. The technique will be useful in predicting distant metastatic sites before distant metastasis occurs, which in turn will help clinicians determine treatment options for cancer patients. © 2025 Bentham Science Publishers.

关键词： Diseases

来源：评论

学校读者我要写书评

暂无评论

Robust Place Recognition Using Fusion of Thermal Infrared and RGB Cameras

引用

Journal of Institute of Control, Robotics and Systems 2024年第12期30卷 1414-1421页

作者： Ma, Seungjun Cho, Younggun Department of Electrical and Computer Engineering Inha University Korea Republic of

The performance of camera-based place recognition has significantly improved with the rapid advancement of deep learning. However, RGB cameras still face challenges in handling variations in lighting conditions due to their inherent limitations. Recent research has explored the use of thermal infrared (TIR) cameras to achieve illumination-invariant scene recognition. Nevertheless, TIR cameras can also struggle in scenarios with uniform heat distribution or when objects with extreme temperatures are present. We herein propose a robust place recognition methodology that leverages RGB-TIR feature fusion. We developed a novel network architecture that can dynamically extract information weights from each RGB and TIR image and then generate a weighted descriptor. This approach is particularly effective in scenarios in which one of the sensor modalities is degraded. The proposed method was validated using a dataset covering a subterranean environment, and it demonstrated superior performance compared with existing visual place recognition (VPR) methods. Additionally, we conducted an ablation study to evaluate the effectiveness of the proposed weight extraction network, providing qualitative insights into its performance. © ICROS 2024.

关键词： SLAM robotics

来源：评论

学校读者我要写书评

暂无评论

Improving the Performance of Universal Traffic Light Recognition Through Dataset Construction and Selection in Semi-supervised Learning

引用

Journal of Institute of Control, Robotics and Systems 2024年第8期30卷 787-792页

作者： Kim, Dayoung Li, Xingyou Kim, Hakil Department of Electrical and Computer Engineering Inha University Korea Republic of

Traffic light recognition in autonomous driving is an essential but very challenging task because its performance is affected by unpredictable environmental conditions. Moreover, the shapes and installations of traffic lights in various countries require a lot of cost and time to collect a huge dataset and train a deep learning model. This study develops a method for constructing training datasets with minimal resource expenditure for more efficient and universally applicable traffic light recognition. It aims to develop a deep learning scheme that can universally recognize various international traffic light configurations by replacing the initial supervised learning stage with a phase that can recognize multiple types of traffic lights and improve performance through self-supervised learning on new datasets. Additionally, this study introduces a data selection algorithm that enables robust recognition under changes in road conditions by enhancing adaptability across diverse environments. This research not only demonstrates improved recognition performance in hazardous road situations but also concludes with potential for efficient application worldwide. Experimental results showed 24.2% improvement in mAP50, even with the same proportion of labeled data, and similar performance with half the labeled data. © ICROS 2024.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

A new multi-input two-output switched-source multilevel inverter based on harvest of maximum energy for photovoltaic applications

引用

International Journal of Power Electronics 2024年第2期19卷 167-192页

作者： Salehi, Seyed Jafar Shamsi-Nejad, Mohammad Ali Najafi, Hamid Reza Department of Electrical and Computer Engineering University of Birjand Birjand Iran

In multilevel inverters, unused energies are created due to the asynchronous use of the input DC-sources. This means that when the input DC-sources are replaced by renewable systems such as photovoltaic arrays, some of the input energies remain unused by the multilevel inverter. This paper presents a new multi-input two output multilevel inverter topology that can harvest the unused energies and return them to another load which leads to the harvest of the maximum input energy. The harvest of maximum energy (HME)-based multilevel inverter structure consists of two independent terminals both of which can be connected to the AC load. Another merit of the proposed multilevel inverter is that the number of its switches is comparable to other structures where unused energies cannot be harvested. The nearest level control (NLC) has been used as the switching strategy in the proposed multilevel structure. In order to verify the performance of the HME-based multilevel inverter topology, the simulation and experimental results for a type seven-level inverter are provided by MATLAB and TMS320F28379D DSP, respectively. © 2024 Inderscience Enterprises Ltd.

关键词： Level control

来源：评论

学校读者我要写书评

暂无评论

Recent progress of artificial intelligence for liquid-vapor phase change heat transfer

引用

npj Computational Materials 2024年第1期10卷 2552-2565页

作者： Youngjoon Suh Aparna Chandramowlishwaran Yoonjin Won Department of Mechanical and Aerospace Engineering University of CaliforniaIrvineCAUSA Department of Electrical Engineering and Computer Science University of CaliforniaIrvineCAUSA

Artificial intelligence(AI)is shifting the paradigm of two-phase heat transfer *** innovations in AI and machine learning uniquely offer the potential for collecting new types of physically meaningful features that have not been addressed in the past,for making their insights available to other domains,and for solving for physical quantities based on first principles for phasechange thermofluidic *** review outlines core ideas of current AI technologies connected to thermal energy science to illustrate how they can be used to push the limit of our knowledge boundaries about boiling and condensation *** technologies for meta-analysis,data extraction,and data stream analysis are described with their potential challenges,opportunities,and alternative ***,we offer outlooks and perspectives regarding physics-centered machine learning,sustainable cyberinfrastructures,and multidisciplinary efforts that will help foster the growing trend of AI for phase-change heat and mass transfer.

关键词： transfer meaningful shifting

来源：评论

学校读者我要写书评

暂无评论

A new content-aware image resizing based on Rényi entropy and deep learning

引用

Neural Computing and Applications 2024年第15期36卷 8885-8899页

作者： Ayubi, Jila Chehel Amirani, Mehdi Valizadeh, Morteza Department of Electrical and Computer Engineering Urmia University Urmia Iran

One of the most popular techniques for changing the purpose of an image or resizing a digital image with content awareness is the seam-carving method. The performance of image resizing algorithms based on seam machining shows that these algorithms are highly dependent on the extraction of importance map techniques and the detection of salient objects. So far, various algorithms have been proposed to extract the importance map. In this paper, a new method based on Rényi entropy is proposed to extract the importance map. Also, a deep learning network has been used to detect salient objects. The simulator results showed that combining Rényi’s importance map with a deep network of salient object detection performed better than classical seam-carving and other extended seam-carving algorithms based on deep learning. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Deep learning

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：