检索结果-内蒙古大学图书馆

SSRN 2024年

作者： Ju, Xiaolin Cao, Yi Chen, Xiang Gong, Lina Chakma, Vaskar Zhou, Xin School of Artificial Intelligence and Computer Science Nantong University Jiangsu Nantong226019 China School of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Jiangsu Nanjing211106 China

Context: Just-in-time defect prediction (JIT-DP) is a crucial process in software development that focuses on identifying potential defects during code changes, facilitating early mitigation and quality assurance. Pre-trained language models like CodeBERT have shown promise in various applications but often struggle to distinguish between defective and non-defective code, especially when dealing with noisy ***: The primary aim of this study is to enhance the robustness of pre-trained language models in identifying software defects by developing an innovative framework that leverages contrastive learning and feature ***: We introduce JIT-CF, a framework that improves model robustness by employing contrastive learning to maximize similarity within positive pairs and minimize it between negative pairs, thereby enhancing the model’s ability to detect subtle differences in code changes. Additionally, feature fusion is used to combine semantic and expert features, enabling the model to capture richer contextual information. This integrated approach aims to improve the identification and resolution of code ***: JIT-CF was evaluated using the JIT-Defects4J dataset, which includes 23,379 code commits from 21 projects. The results indicate substantial performance improvements over seven state-of-the-art baselines, with enhancements of up to 13.9% in F1-score, 8% in AUC, and 11% in Recall@20%E. The study also explores the impact of specific customization enhancements, demonstrating the potential for improved just-in-time defect ***: The proposed JIT-CF framework significantly advances the field of just-in-time defect prediction by effectively addressing the challenges encountered by pre-trained models in distinguishing code defects. The integration of contrastive learning and feature fusion not only enhances the model’s robustness but also leads to notable improvements in prediction accuracy, offering valuable insights for

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

EmoDamp: A Model for Speaker Emotion Inference in Dialogues with Bidirectional Emotion Damping Mechanism 6

EmoDamp: A Model for Speaker Emotion Inference in Dialogues ...

引用

6th International Conference on Algorithms, Computing and artificial intelligence, ACAI 2023

作者： Jiang, Yueqi Sun, Xiao Wang, Jiamin AnHui Province Key Laboratory of Affective Computing and Advanced Intelligent Machine School of Computer Science and Information Engineering Hefei University of Technology Anhui Hefei China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Anhui Hefei China

ISBN: (纸本)9798400709203

Emotion analysis in conversation has been a popular research topic in the natural language processing field. While much of the existing research has focused on emotion recognition in conversation, the emotion inference task in conversation is more challenging due to its specific prerequisites. In this paper, we propose EmoDamp, a novel model for the conversational emotion inference task. By referring to psychological theories, a bidirectional emotion damping network(BED Net) is designed to model the emotion generation mechanism. Furthermore, external commonsense knowledge is also used to supplement the prediction results to make them more consistent with authentic logic. Finally, a conditional random field (CRF) is used to capture the potential constraints of the emotion transfer process in conversation and simulate emotion contagion. Experimental results on two benchmark datasets demonstrate that EmoDamp outperforms the baseline model, achieving improved results on each emotion category and showcasing its effectiveness. © 2023 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Human computer interaction

来源：评论

学校读者我要写书评

暂无评论

POCE: Pose-Controllable Expression Editing

引用

IEEE Transactions on Image Processing 2023年 32卷 6210-6222页

作者： Wu, Rongliang Yu, Yingchen Zhan, Fangneng Zhang, Jiahui Liao, Shengcai Lu, Shijian Nanyang Technological University School of Computer Science and Engineering Singapore639798 Singapore Singapore138632 Singapore Max Planck Max Planck Institute for Informatics Saarbrücken66123 Germany Inception Institute of Artificial Intelligence Abu Dhabi United Arab Emirates

Facial expression editing has attracted increasing attention with the advance of deep neural networks in recent years. However, most existing methods suffer from compromised editing fidelity and limited usability as they either ignore pose variations (unrealistic editing) or require paired training data (not easy to collect) for pose controls. This paper presents POCE, an innovative pose-controllable expression editing network that can generate realistic facial expressions and head poses simultaneously with just unpaired training images. POCE achieves the more accessible and realistic pose-controllable expression editing by mapping face images into UV space, where facial expressions and head poses can be disentangled and edited separately. POCE has two novel designs. The first is self-supervised UV completion that allows to complete UV maps sampled under different head poses, which often suffer from self-occlusions and missing facial texture. The second is weakly-supervised UV editing that allows to generate new facial expressions with minimal modification of facial identity, where the synthesized expression could be controlled by either an expression label or directly transplanted from a reference UV map via feature transfer. Extensive experiments show that POCE can learn from unpaired face images effectively, and the learned model can generate realistic and high-fidelity facial expressions under various new poses. © 1992-2012 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

data Augmentation for Detection of Bacteria from Gram Stained Smears Images by YOLOv8 16

Data Augmentation for Detection of Bacteria from Gram Staine...

引用

16th IIAI International Congress on Advanced Applied Informatics, IIAI-AAI 2024

作者： Nakajo, Ryunosuke Tsutsui, Minori Hirata, Kouichi Graduate School of Computer Science and Systems Engineering Kyushu Institute of Technology Kawazu 680-4 Iizuka820-8502 Japan Kyushu Institute of Technology Department of Artificial Intelligence Kawazu 680-4 Iizuka820-8502 Japan Ns Solutions Kyushu Corporation Japan

ISBN: (纸本)9798350377903

In this paper, we investigate the methods of data augmentation appropriate to detect bacteria from Gram stained smears images by YOLOv8. For data augmentation, we adopt the flipping, the rotation, the cutout and the change of the values in HSB color space. Then, we evaluate these methods and the methods of the cutout and the change of the values in HSB color space after applying the flipping and the rotation. © 2024 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for Automated Classification of Kidney Lesions in Chronic Kidney Disease Patients 2

Deep Learning for Automated Classification of Kidney Lesions...

引用

2nd IEEE International Conference on Integrated Circuits and Communication Systems, ICICACS 2024

作者： Lakshmi, B. Seetha Anand, Vijay Omotunde, Habeeb Birundadevi, M. Mahalakshmi, S. Brahmam, Madala Guru Sri Shanmugha College of Engineering and Technology Department of Artificial Intelligence and Data Science Tamilnadu Namakkal India School of Computer Science Engineering & Information Systems Vellore Institute of Technology Vellore India College of Computer and Information Sciences Imam Mohammad Ibn Saud Islamic University Department of Information Systems Riyadh Saudi Arabia Sri Sairam Siddha Medical College and Research Centre Department of Maruthuvam Chennai India R.M.K. Engineering College Department of Computer Science and Design Tamilnadu Chennai India

ISBN: (纸本)9798350317558

Deep Learning was utilized to autonomously classify renal abnormalities, which improved the diagnostic accuracy for individuals with chronic kidney disease (CKD), according to this study. During training on a meticulously selected dataset, the proposed model outperformed the most recent approaches. By utilizing medical imaging data - specifically quantitative and qualitative data - to differentiate between discrete kidney abnormalities, the model demonstrated exceptional levels of accuracy, precision, and recall. Furthermore, apart from effectively resolving ethical considerations pertaining to data privacy, this innovative methodology has the potential to fundamentally transform the existing landscape of chronic kidney disease (CKD) diagnosis and treatment. The results demonstrate that the model effectively manages intricate renal disorders, signifying a noteworthy progression in the field of medical image processing. This research paves the way for additional fruitful lines of inquiry, which ought to enhance nephrology diagnostics and patient care. © 2024 IEEE.

关键词： Automated Classification Chronic Kidney Disease Deep Learning Diagnostic Accuracy Kidney Lesions Medical Image Analysis

来源：评论

学校读者我要写书评

暂无评论

data Mining for Predictive Analytics and Optimization of Treatment Plans in Cardiovascular Disease Management using Neural Networks 2

Data Mining for Predictive Analytics and Optimization of Tre...

引用

2nd International Conference on Advancements in Smart, Secure and Intelligent Computing, ASSIC 2024

作者： Kaliappan, Seeniappan Balaji, V. Bharathi, G. Bhavani Aluvala, Srinivas Kcg College of Technology Department of Mechatronics Engineering Karapakkam Tamil Nadu Chennai600097 India Loyola Institute of Technology Department of Mechanical Engineering Tamil Nadu Chennai600123 India School of Computer Science & Artificial Intelligence Sr University Telangana Warangal506371 India

ISBN: (纸本)9798350370188

This section presents an integrated method that merges data mining techniques and neural networks to augment predictive analytics and optimize treatment strategies for managing cardiovascular diseases. Algorithm 1 focuses on data preprocessing and feature selection, essential for refining the patient data. It standardizes the input data by normalization and selects critical features using information gain, enhancing the subsequent neural network's efficiency in predicting disease progression. Algorithm elucidates the training of a neural network, utilizing the pre-processed data and disease progression labels. The neural network undergoes iterative training to minimize the loss function, resulting in a trained model capable of accurate predictions. Algorithm emphasizes the utilization of the trained neural network to optimize personalized treatment plans for cardiovascular disease management. By predicting disease progression and evaluating treatment options, it tailors' treatment plans to individual patient needs, thereby enhancing the effectiveness of healthcare interventions. The proposed method is demonstrated to surpass traditional techniques such as Linear Regression, Decision Trees, Support Vector Machines, K-Nearest Neighbors, Random Forest, and Logistic Regression in terms of critical performance metrics. Leveraging the power of neural networks, the proposed method showcases superior accuracy, precision, recall, Fl score, and AUC-ROC. Moreover, a comprehensive evaluation involving sensitivity, specificity, Matthews Correlation Coefficient (MCC), Negative Predictive Value (NPV), Youden's Index, and Cohen's Kappa consistently emphasizes its advanced performance, validating its potential to revolutionize cardiovascular disease management. © 2024 IEEE.

关键词： data mining

来源：评论

学校读者我要写书评

暂无评论

A Collaborative Filtering Algorithm with Improved User Similarity Calculation for Enhanced Recommendation Accuracy and Personalization 12

A Collaborative Filtering Algorithm with Improved User Simil...

引用

12th International Conference of Information and Communication technology, ICTech 2023

作者： Ding, Shili Fan, Yue Liu, Jing Zhang, Yonghong Shenyang Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China School of Information Science and Technology Taishan University Shandong Taian China Shandong University Laboratory of Big Data and Artificial Intelligence Technology Jinan China

ISBN: (纸本)9798350332919

The conventional method for calculating user similarity in the collaborative filtering recommendation algorithm based on users fails to address the challenge of variations in the number, value, and popularity of items that are collaboratively rated by multiple users. In order to solve the above problems, a new user similarity calculation method is proposed. By fusing cosine similarity and Jackard similarity, the problem of difference in the number of common scoring items is alleviated, so as to reduce the error of scoring prediction. Finally, experiments were performed on the MovieLens dataset, and the findings demonstrated that the enhanced algorithm outperforms the established benchmark similarity algorithm regarding mean absolute error (MAE). In addition, the improved algorithm has good robustness in predicting the accuracy of user ratings and recommendation quality. © 2023 IEEE.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

LR-IAD:Mask-Free Industrial Anomaly Detection with Logical Reasoning

arXiv

引用

arXiv 2025年

作者： Zeng, Peijian Pang, Feiyan Wang, Zhanbo Yang, Aimin School of Computer Science and Technology Guangdong University of Technology Guangzhou China School of Computer Science and Intelligence Education Lingnan Normal University Zhanjian China

Industrial Anomaly Detection (IAD) is critical for ensuring product quality by identifying defects. Traditional methods such as feature embedding and reconstruction-based approaches require large datasets and struggle with scalability. Existing vision-language models (VLMs) and Multimodal Large Language Models (MLLMs) address some limitations but rely on mask annotations, leading to high implementation costs and false positives. Additionally, industrial datasets like MVTec-AD and VisA suffer from severe class imbalance, with defect samples constituting only 23.8% and 11.1% of total data respectively. To address these challenges, we propose a reward function that dynamically prioritizes rare defect patterns during training to handle class imbalance. We also introduce a mask-free reasoning framework using Chain of Thought (CoT) and Group Relative Policy Optimization (GRPO) mechanisms, enabling anomaly detection directly from raw images without annotated masks. This approach generates interpretable step-by-step explanations for defect localization. Our method achieves state-of-the-art performance, outperforming prior approaches by 36% in accuracy on MVTec-AD and 16% on VisA. By eliminating mask dependency and reducing costs while providing explainable outputs, this work advances industrial anomaly detection and supports scalable quality control in manufacturing. Code to reproduce the experiment is available at https://***/LilaKen/LR-IAD. Copyright © 2025, The Authors. All rights reserved.

关键词： Damage detection

来源：评论

学校读者我要写书评

暂无评论

Enhancing adaptive history reserving by spiking convolutional block attention module in recurrent neural networks 23

Enhancing adaptive history reserving by spiking convolutiona...

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Qi Xu Yuyuan Gao Jiangrong Shen Yaxin Li Xuming Ran Huajin Tang Gang Pan School of Artificial Intelligence Dalian University of Technology College of Computer Science and Technology Zhejiang University

Spiking neural networks (SNNs) serve as one type of efficient model to process spatio-temporal patterns in time series, such as the Address-Event Representation data collected from Dynamic Vision Sensor (DVS). Although convolutional SNNs have achieved remarkable performance on these AER datasets, benefiting from the predominant spatial feature extraction ability of convolutional structure, they ignore temporal features related to sequential time points. In this paper, we develop a recurrent spiking neural network (RSNN) model embedded with an advanced spiking convolutional block attention module (SCBAM) component to combine both spatial and temporal features of spatio-temporal patterns. It invokes the history information in spatial and temporal channels adaptively through SCBAM, which brings the advantages of efficient memory calling and history redundancy elimination. The performance of our model was evaluated in DVS128-Gesture dataset and other time-series datasets. The experimental results show that the proposed SRNN-SCBAM model makes better use of the history information in spatial and temporal dimensions with less memory space, and achieves higher accuracy compared to other models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Arbitrary Clothing Style Transfer Based on Attention Mechanism 27th

Arbitrary Clothing Style Transfer Based on Attention Mechan...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Huang, Chen Zhang, Junjie Yuan, Hua School of Computer Science and Artificial Intelligence Wuhan Textile University Wuhan430200 China Hubei Key Laboratory of Digital Textile Equipment Wuhan Textile University Wuhan430073 China Wuhan Textile and Apparel Digital Engineering Technology Research Center School of Fashion Wuhan Textile University Wuhan430073 China

ISBN: (纸本)9783031781247

In recent years, style transfer has become increasingly prominent in various domains, especially fashion. As a tool for designers, clothing style transfer generates a wide array of styles, enabling rapid experimentation and fostering creative inspiration. However, current methods have poor performance in transferring textures and colors from style images to clothing, and commonly result in blurred boundaries between clothing and background. To address these challenges, an arbitrary style transfer algorithm for clothing is proposed, utilizing attention network and feature fusion for more effective and efficient style application. In this paper, the criss-cross attention network is incorporated to extract comprehensively global features and capture long-range dependencies, thus minimizing artifacts and enhancing texture transfer. Through a novel multi-level feature fusion approach, color transfer becomes more natural and coherent, closely aligning with the color palette of the style image. Additionally, semantic segmentation is employed to separate clothing from the background, preserving the original background and character. The experimental results show that the user preference of this paper’s algorithm is much higher than existing methods, and single 512×512 image style transfer takes only 0.314 s with real-time performance. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Semantic Segmentation

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：