检索结果-内蒙古大学图书馆

12th International Conference on Learning Representations, ICLR 2024

作者： Patel, Yash J. Kundu, Akash Ostaszewski, Mateusz Bonet-Monroig, Xavier Dunjko, Vedran Danaci, Onur 〈aQaL〉 Applied Quantum Algorithms Leiden University Netherlands Institute of Theoretical and Applied Informatics Polish Academy of Sciences Poland Joint Doctoral School Silesian University of Technology Poland Warsaw University of Technology Institute of Computer Science Poland Lorentz Institute Leiden University Netherlands LIACS Leiden University Netherlands

The key challenge in the noisy intermediate-scale quantum era is finding useful circuits compatible with current device limitations. Variational quantum algorithms (VQAs) offer a potential solution by fixing the circuit architecture and optimizing individual gate parameters in an external loop. However, parameter optimization can become intractable, and the overall performance of the algorithm depends heavily on the initially chosen circuit architecture. Several quantum architecture search (QAS) algorithms have been developed to design useful circuit architectures automatically. In the case of parameter optimization alone, noise effects have been observed to dramatically influence the performance of the optimizer and final outcomes, which is a key line of study. However, the effects of noise on the architecture search, which could be just as critical, are poorly understood. This work addresses this gap by introducing a curriculum-based reinforcement learning QAS (CRLQAS) algorithm designed to tackle challenges in realistic VQA deployment. The algorithm incorporates (i) a 3D architecture encoding and restrictions on environment dynamics to explore the search space of possible circuits efficiently, (ii) an episode halting scheme to steer the agent to find shorter circuits, and (iii) a novel variant of simultaneous perturbation stochastic approximation as an optimizer for faster convergence. To facilitate studies, we developed an optimized simulator for our algorithm, significantly improving computational efficiency in simulating noisy quantum circuits by employing the Pauli-transfer matrix formalism in the Pauli-Liouville basis. Numerical experiments focusing on quantum chemistry tasks demonstrate that CRLQAS outperforms existing QAS algorithms across several metrics in both noiseless and noisy environments. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Inpainting applied to Facade Images: A Comparison of Algorithms 3rd

Inpainting Applied to Facade Images: A Comparison of Algorit...

引用

3rd International Conference on Pattern Recognition and Artificial Intelligence, ICPRAI 2022

作者： Fritzsche, Willy Goebbels, Steffen Hensel, Simon Rußinski, Marco Schuch, Nils Faculty of Electrical Engineering and Computer Science iPattern Institute Niederrhein University of Applied Sciences Krefeld47805 Germany

ISBN: (纸本)9783031090363

Many municipalities provide textured 3D city models for planning and simulation purposes. Usually, the textures are automatically taken from oblique aerial images. In these images, walls may be occluded by building parts, vegetation and other objects such as cars, traffic signs, etc. To obtain high quality models, these objects have to be segmented and then removed from facade textures. In this study, we investigate the ability of different non-specialized inpainting algorithms to continue facade patterns in occluded facade areas. To this end, non-occluded facade textures of a 3D city model are equipped with various masks simulating occlusions. Then, the performance of the algorithms is evaluated by comparing their results with the original images. In particular, very useful results are obtained with the neural network "DeepFill v2" trained with transfer learning on freely available facade datasets and the "Shift-Map" algorithm. © 2022, Springer Nature Switzerland AG.

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

SynDronevision: A Synthetic Dataset for Image-Based Drone Detection

SynDroneVision: A Synthetic Dataset for Image-Based Drone De...

引用

IEEE Workshop on Applications of computer vision (WACV)

作者： Tamara R. Lenhard Andreas Weinmann Kai Franke Tobias Koch Institute for the Protection of Terrestrial Infrastructures German Aerospace Center (DLR) Germany Working Group Algorithms for Computer Vision Imaging and Data Analysis University of Applied Sciences Darmstadt Germany

ISBN: (数字)9798331510831

ISBN: (纸本)9798331510848

Developing robust drone detection systems is often constrained by the limited availability of large-scale annotated training data and the high costs associated with real-world data collection. However, leveraging synthetic data generated via game engine-based simulations provides a promising and cost-effective solution to overcome this issue. Therefore, we present SynDronevision, a synthetic dataset specifically designed for RGB-based drone detection in surveillance applications. Featuring diverse backgrounds, lighting conditions, and drone models, SynDronevision offers a comprehensive training foundation for deep learning algorithms. To evaluate the dataset's effectiveness, we perform a comparative analysis across a selection of recent YOLO detection models. Our findings demonstrate that SynDronevision is a valuable resource for real-world data enrichment, achieving notable enhancements in model performance and robustness, while signifcantly reducing the time and costs of real-world data acquisition. SynDronevision can be accessed at https://***/records/13360116.

关键词： YOLO Training Costs Surveillance Training data Lighting Data models Robustness Drones Synthetic data

来源：评论

学校读者我要写书评

暂无评论

Confidence-calibrated covariate shift correction for few-shot classification in vision-Language Models

arXiv

引用

arXiv 2025年

作者： Khan, Behraj Qureshi, Rizwan Muhammad Durrani, Nouman Qasim Syed, Tahir School of Mathematics and Computer Science Institute of Business Administration Karachi Pakistan Center for Research in Computer Vision University of Central Florida United States National University of Computer and Emerging Sciences Karachi Pakistan

Since the establishment of vision-language foundation models as the new mainstay in low-shot vision classification tasks, the question of domain generalization arising from insufficient target data is assuming more importance. This scarcity challenge induces sampling bias and amplifies model sensitivity to variations and shifts in data distributions. While fine-tuning on multiple domains could mitigate such domain generalization issues, it is resource-intensive and demands diverse data sources. In this work, we systematically analyze two critical challenges: (1) covariate shift between the pre-training distribution and the underspecified target distribution, and (2) confidence misalignment, where predictions on novel data are overconfident. To address both challenges simultaneously, we introduce Confidence-Calibrated Covariate Shift Correction (CalShift)—a unified approach that combines a Fisher information penalty to mitigate covariate shift and a Confidence Misalignment Penalty (CMP) to reduce overconfidence in misclassified examples. Experimental evaluations across various vision and covariate shift benchmarks demonstrate that CalShift significantly improves model calibration, achieving up to a 5.82% reduction in Expected Calibration Error (ECE). Furthermore, CalShift enhances robustness, improving accuracy by 3.5% on challenging datasets impacted by covariate shifts. Our results highlight CalShift as a promising strategy for building robust and reliable low-shot vision-language systems for real-world applications. © 2025, CC BY-NC-SA.

关键词： Fisher information matrix

来源：评论

学校读者我要写书评

暂无评论

Investigation of metabolic pathways from gut microbiome analyses regarding type 2 diabetes mellitus using artificial neural networks

引用

Discover Artificial Intelligence 2023年第1期3卷 19页

作者： Siptroth, Julienne Moskalenko, Olga Krumbiegel, Carsten Ackermann, Jörg Koch, Ina Pospisil, Heike High Performance Computing in Life Sciences Technical University of Applied Sciences Wildau Wildau15745 Germany BIOMES NGS GmbH Schwartzkopfstraße 1 Wildau15745 Germany Department of Molecular Bioinformatics Institute of Computer Science Goethe University Frankfurt Frankfurt am Main60325 Germany

Background: Type 2 diabetes mellitus is a prevalent disease that contributes to the development of various health issues, including kidney failure and strokes. As a result, it poses a significant challenge to the worldwide healthcare system. Research into the gut microbiome has enabled the identification and description of various diseases, with bacterial pathways playing a critical role in this context. These pathways link individual bacteria based on their biological functions. This study deals with the classification of microbiome pathway profiles of type 2 diabetes mellitus patients. Methods: Pathway profiles were determined by next-generation sequencing of 16S rDNA from stool samples, which were subsequently assigned to bacteria. Then, the involved pathways were assigned by the identified gene families. The classification of type 2 diabetes mellitus is enabled by a constructed neural network. Furthermore, a feature importance analysis was performed via a game theoretic approach (SHapley Additive exPlanations). The study not only focuses on the classification using neural networks, but also on identifying crucial bacterial pathways. Results: It could be shown that a neural network classification of type 2 diabetes mellitus and a healthy comparison group is possible with an excellent prediction accuracy. It was possible to create a ranking to identify the pathways that have a high impact on the model prediction accuracy. In this way, new associations between the alteration of, e.g. a biosynthetic pathway and the presence of diabetes mellitus type 2 disease can also be discovered. The basis is formed by 946 microbiome pathway profiles from diabetes mellitus type 2 patients (272) and healthy comparison persons (674). Conclusion: With this study of the gut microbiome, we present an approach using a neural network to obtain a classification of healthy and type 2 diabetes mellitus and to identify the critical features. Intestinal bacteria pathway profiles form the bas

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

A Hardware-Aware Network for Real-World Single Image Super-Resolutions

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第7期5卷 3482-3496页

作者： Ma, Rui Du, Xian Institute for Applied Life Sciences University of Massachusetts Center for Personalized Health Monitoring AmherstMA01003 United States University of Massachusetts Electrical and Computer Engineering Department AmherstMA01003 United States University of Massachusetts Mechanical and Industrial Engineering Department AmherstMA01003 United States

Most single image super-resolution (SISR) methods are developed on synthetic low-resolution (LR) and high-resolution (HR) image pairs, which are simulated by a predetermined degradation operation, such as bicubic downsampling. However, these methods only learn the inverse process of the predetermined operation, which fails to super resolve the real-world LR images, whose true formulation deviates from the predetermined operation. To address this, we propose a novel super-resolution (SR) framework named hardware-aware super-resolution (HASR) network that first extracts hardware information, particularly the camera degradation information. The LR images are then super resolved by integrating the extracted information. To evaluate the performance of HASR network, we build a dataset named Real-Micron from real-world micron-scale patterns. The paired LR and HR images are captured by changing the objectives and registered using a developed registration algorithm. Transfer learning is implemented during the training of Real-Micron dataset due to the lack of amount of data. Experiments demonstrate that by integrating the degradation information, our proposed network achieves state-of-the-art performance for the blind SR task on both synthetic and real-world datasets. © 2020 IEEE.

关键词： Optical resolving power

来源：评论

学校读者我要写书评

暂无评论

The Impact of Augmentation and SMOTE Implementation on the Classification Models Performance: A Case Study on Student Academic Performance Dataset 9th

The Impact of Augmentation and SMOTE Implementation on the C...

引用

9th International Conference on computer Science and Computational Intelligence, ICCSCI 2024

作者： Sano, Albert Verasius Dian Bhatti, Faqir M. Miranda, Eka Aryuni, Mediana Zakiyyah, Alfi Yusrotis Bernando, Charles Computer Science Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Riphah Institute of Computing and Applied Sciences Riphah Intternational University Raiwind Lahore Pakistan Information Systems Department School of Information Systems Bina Nusantara University Jakarta11480 Indonesia Mathematics Department School of computer Science Bina Nusantara University Jakarta11480 Indonesia

This study aims to find out the impact of data augmentation and Synthetic Minority Over-sampling Techniques (SMOTE) implementation on the classification models performance using small and imbalanced dataset of student academic performance. The design of this study involved a comprehensive experiment by comparing four scenarios: 1) comparing classification models without both data augmentation and SMOTE, 2) models with data augmentation, 3) models with SMOTE, and 4) models with both data augmentation and SMOTE. The model's performances were each measured based on standard evaluation metrices such as accuracy, precision, recall, F1-score. To test the results validity, there were three classification algorithms implemented and evaluated for each scenario, that is, Random Forest, XGBoost, and AdaBoost. The finding of this study highlights the significant impact of data augmentation and SMOTE to the increase of classification models performance, particularly over the small and imbalanced dataset. Results showed that the implementation of both techniques simultaneously brought about the most significant increase in the evaluation metrices compared to the implementation of both techniques separately. The originality of this study lies in its comprehensive approach in comparing the effectiveness of data augmentation and SMOTE, as well as the use of student academic performance dataset, which is, a real case in the context of artificial intelligence. This finding gives a valuable insight to the researchers and practitioners in choosing appropriate techniques to handle small and imbalanced class datasets. This study is expected to make an important contribution to the more effective development of classification methodology in various domains. © 2024 The Authors.

关键词： Data accuracy

来源：评论

学校读者我要写书评

暂无评论

Alzheimer's Disease Diagnosis and Categorization using Deep Learning 1

Alzheimer's Disease Diagnosis and Categorization using Deep ...

引用

1st IEEE International Conference on Cognitive Robotics and Intelligent Systems, ICC - ROBINS 2024

作者： Sangeetha, M. Gunasekaran, Hemalatha Azalini, T. Nakshathra, J. Uma Mageshwari, P. Manjula Devi, R. Dr.N.G.P.Institute of Technology Department of Information Technology Tamilnadu Coimbatore India University of Technology and Applied Sciences College of Computing and Information Sciences Ibri Oman Kpr Institute of Engineering and Technology Department of Computer Science and Engineering Tamilnadu Coimbatore India

ISBN: (纸本)9798350372748

Alzheimer's disease (AD) is a significant global health concern, accounting for the sixth highest cause of mortality worldwide. Early diagnosis is critical to improving patient care and treatment outcomes. This study aims to classify Alzheimer's disease (AD) using Magnetic Resonance Imaging (MRI) brain images, allowing for prompt intervention and optimal disease treatment. The dataset consists of 6400 MRI pictures classified into four categories: very mildly demented, mildly demented, moderately demented, and non-demented, with an 80-10-10 split for training, validation, and testing. Initially, the utilization of VGG16 architecture results with an accuracy of 87%. To improve accuracy and reduce training time, the GoogleNet (Inception V3) Convolutional Neural Network (CNN) architecture was used. This method resulted with an accuracy of 97.62%. © 2024 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

UAV swarm neural network: Infrastructure of relevant telecommunication systems 6

UAV swarm neural network: Infrastructure of relevant telecom...

引用

6th IEEE International Conference on Information and Telecommunication Technologies and Radio Electronics, UkrMiCo 2023

作者： Pechurin, Nickolay Kondratova, Lyudmila National Aviation University Department of Computer Sciences and Technologies Kyiv Ukraine National Technical University of Ukraine Igor Sikorsky Kyiv Polytechnic Institute Institute for Applied System Analysis Kyiv Ukraine

ISBN: (纸本)9798350348484

In the article it is compared the architecture options of artificial neural networks, the elements of which are placed on unmanned aerial vehicles, in particular light and ultralight, and the physical infrastructure of the corresponding telecommunication systems. Architectures of single-layer and multi-layer forward-directed, recurrent and fully connected networks are considered. The given comparison will make it possible to use simple adequate physical structures of telecommunications systems and appropriate information technologies in the creation of onboard neural networks and neurocomputing systems. It will also allow determining the physical parameters of network architectures adapted to operating conditions in an adverse environment. As an example, a variant of the telecommunications system architecture for placing the simplest swarm neurocomputer for pattern recognition is given. © 2023 IEEE.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

LogicAD: Explainable Anomaly Detection via VLM-based Text Feature Extraction 39

LogicAD: Explainable Anomaly Detection via VLM-based Text Fe...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Jin, Er Feng, Qihui Mou, Yongli Lakemeyer, Gerhard Decker, Stefan Simons, Oliver Stegmaier, Johannes Institute of Imaging and Computer Vision RWTH Aachen University Aachen Germany Department of Computer Science RWTH Aachen University Aachen Germany Fraunhofer Institute for Applied Information Technology FIT Sankt Augustin Germany

ISBN: (纸本)157735897X

Logical image understanding involves interpreting and reasoning about the relationships and consistency within an image’s visual content. This capability is essential in applications such as industrial inspection, where logical anomaly detection (AD) is critical for maintaining high-quality standards and minimizing costly recalls. Previous research in AD has relied on prior knowledge for designing algorithms, which often requires extensive manual annotation effort, significant computing power, and large amounts of data for training. Autoregressive, multimodal vision Language Models (AVLMs) offer a promising alternative due to their exceptional performance in visual reasoning across various domains. Despite this, their application in logical AD remains unexplored. In this work, we investigate using AVLMs for logical AD and demonstrate that they are well-suited to the task. Combining AVLMs with format embedding and a logic reasoner, we achieve state-of-the-art (SOTA) AD performance on public benchmarks, MVTec LOCO AD, with an AUROC of 86.0% and an F1-max of 83.7% along with explanations of the anomalies. This significantly outperforms the existing SOTA method by 18.1% in AUROC and 4.6% in F1-max score. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Anomaly 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：