检索结果-内蒙古大学图书馆

40th International Conference on Machine Learning, ICML 2023

作者： Yan, Zhiqiang Li, Xiang Wang, Kun Chen, Shuo Li, Jun Yang, Jian PCALab School of Computer Science and Engineering Nanjing University of Science and Technology China RIKEN Center for Advanced Intelligence Project Japan

Standard MSE or MAE loss function is commonly used in limited field-of-vision depth completion, treating each pixel equally under a basic assumption that all pixels have same contribution during optimization. Recently, with the rapid rise of panoramic photography, panoramic depth completion (PDC) has raised increasing attention in 3D computer vision. However, the assumption is inapplicable to panoramic data due to its latitudewise distortion and high uncertainty nearby textures and edges. To handle these challenges, we propose distortion and uncertainty aware loss (DUL) that consists of a distortion-aware loss and an uncertainty-aware loss. The distortionaware loss is designed to tackle the panoramic distortion caused by equirectangular projection, whose coordinate transformation relation is used to adaptively calculate the weight of the latitudewise distortion, distributing uneven importance instead of the equal treatment for each pixel. The uncertainty-aware loss is presented to handle the inaccuracy in non-smooth regions. Specifically, we characterize uncertainty into PDC solutions under Bayesian deep learning framework, where a novel consistent uncertainty estimation constraint is designed to learn the consistency between multiple uncertainty maps of a single panorama. This consistency constraint allows model to produce more precise uncertainty estimation that is robust to feature deformation. Extensive experiments show the superiority of our method over standard loss functions, reaching the state of the art. © 2023 Proceedings of Machine Learning Research. All rights reserved.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

SfPUEL: Shape from Polarization under Unknown Environment Light 38

SfPUEL: Shape from Polarization under Unknown Environment Li...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Lyu, Youwei Guo, Heng Zhang, Kailong Li, Si Shi, Boxin School of Artificial Intelligence Beijing University of Posts and Telecommunications China State Key Laboratory for Multimedia Information Processing School of Computer Science Peking University China National Engineering Research Center of Visual Technology School of Computer Science Peking University China

Shape from polarization (SfP) benefits from advancements like polarization cameras for single-shot normal estimation, but its performance heavily relies on light conditions. This paper proposes SfPUEL, an end-to-end SfP method to jointly estimate surface normal and material under unknown environment light. To handle this challenging light condition, we design a transformer-based framework for enhancing the perception of global context features. We further propose to integrate photometric stereo (PS) priors from pretrained models to enrich extracted features for high-quality normal predictions. As metallic and dielectric materials exhibit different BRDFs, SfPUEL additionally predicts dielectric and metallic material segmentation to further boost performance. Experimental results on synthetic and our collected real-world dataset demonstrate that SfPUEL significantly outperforms existing SfP and single-shot normal estimation methods. The code and dataset is available at https://***/YouweiLyu/SfPUEL. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

VIPNet: Combining Viewpoint Information and Shape Priors for Instant Multi-view 3D Reconstruction 17th

VIPNet: Combining Viewpoint Information and Shape Priors fo...

引用

17th Asian Conference on computer Vision, ACCV 2024

作者： Ye, Weining Li, Zhixuan Jiang, Tingting School of Computer Science Peking University Beijing100871 China College of Computing and Data Science Nanyang Technological University Singapore Singapore National Engineering Research Center of Visual Technology National Key Laboratory for Multimedia Information Processing School of Computer Science National Biomedical Imaging Center Peking University Beijing100871 China

ISBN: (纸本)9789819609680

While the multi-view 3D reconstruction task has made significant progress, existing methods simply fuse multi-view image features without effectively leveraging available auxiliary information, especially the viewpoint information for guiding and associating features of different views. To this end, we propose to enhance multi-view 3D reconstruction with the power of viewpoint information. Specifically, a simple-yet-effective viewpoint estimator is designed to learn and provide comprehensive viewpoint knowledge for locating and associating learned features from different views. Moreover, to improve the 3D reconstruction quality when 2D images of only very few viewpoints are available, we propose to learn the shape prior knowledge to provide sufficient shape information for compensating the limited 2D observations. Overall, we present VIPNet, benefiting from Viewpoint Information and Shape Prior learning for high-quality multi-view 3D reconstruction. Extensive experiments validate the effectiveness of the proposed VIPNet, which achieves state-of-the-art performance on challenging datasets and shows well generalization ability in real-world scenarios. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： 3D modeling

来源：评论

学校读者我要写书评

暂无评论

Proposed Biometric Security System Based on Deep Learning and Chaos Algorithms

引用

computers, Materials & Continua 2023年第2期74卷 3515-3537页

作者： Iman Almomani Walid El-Shafai Aala AlKhayer Albandari Alsumayt Sumayh S.Aljameel Khalid Alissa Security Engineering Lab Computer Science DepartmentPrince Sultan UniversityRiyadh11586Saudi Arabia Computer Science Department King Abdullah II School of Information TechnologyThe University of Jordan11942Jordan Department of Electronics and Electrical Communications Engineering Faculty of Electronic EngineeringMenoufia UniversityMenouf32952Egypt Computer Science Department Applied CollegeImam Abdulrahman Bin Faisal UniversityP.O.Box 1982Dammam31441Saudi Arabia Computer Science Department College of Computer Science and Information TechnologyImam Abdulrahman Bin Faisal UniversityP.O.Box 1982Dammam31441Saudi Arabia SAUDI ARAMCO Cybersecurity Chair Networks and Communications DepartmentCollege of Computer Science and Information TechnologyImam Abdulrahman Bin Faisal UniversityP.O.Box 1982Dammam31441Saudi Arabia

Nowadays,there is tremendous growth in biometric authentication and cybersecurity ***,the efficient way of storing and securing personal biometric patterns is mandatory in most governmental and private ***,designing and implementing robust security algorithms for users’biometrics is still a hot research area to be *** work presents a powerful biometric security system(BSS)to protect different biometric modalities such as faces,iris,and *** proposed BSSmodel is based on hybridizing auto-encoder(AE)network and a chaos-based ciphering algorithm to cipher the details of the stored biometric patterns and ensures their *** employed AE network is unsupervised deep learning(DL)structure used in the proposed BSS model to extract main biometric *** obtained features are utilized to generate two random chaos *** first random chaos matrix is used to permute the pixels of biometric *** contrast,the second random matrix is used to further cipher and confuse the resulting permuted biometric pixels using a two-dimensional(2D)chaotic logisticmap(CLM)*** assess the efficiency of the proposed BSS,(1)different standardized color and grayscale images of the examined fingerprint,faces,and iris biometrics were used(2)comprehensive security and recognition evaluation metrics were *** assessment results have proven the authentication and robustness superiority of the proposed BSSmodel compared to other existing *** example,the proposed BSS succeeds in getting a high area under the receiver operating characteristic(AROC)value that reached 99.97%and low rates of 0.00137,0.00148,and 3516 CMC,2023,vol.74,no.20.00157 for equal error rate(EER),false reject rate(FRR),and a false accept rate(FAR),respectively.

关键词： Biometric security deep learning AE network 2D CLM cybersecurity and authentication applications feature extraction unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Visual-Language Collaborative Representation Network for Broad-Domain Few-Shot Image Classification 24

Visual-Language Collaborative Representation Network for Bro...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Guo, Qianyu Ren, Jieji Wang, Haofen Wu, Tianxing Ge, Weifeng Zhang, Wenqiang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China School of Mechanical Engineering Shanghai Jiao Tong University Shanghai China College of Design and Innovation Tongji University Shanghai China School of Computer Science and Engineering Southeast University Nanjing China School of Computer Science Fudan University Shanghai China Engineering Research Center of AI & Robotics Ministry of Education Academy for Engineering & Technology Fudan University Shanghai China

ISBN: (纸本)9798400706868

Visual-language models based on CLIP have shown remarkable abilities in general few-shot image classification. However, their performance drops in specialized fields such as healthcare or agriculture, because CLIP's pre-training does not cover all category data. Existing methods excessively depend on the multi-modal information representation and alignment capabilities acquired from CLIP pre-training, which hinders accurate generalization to unfamiliar domains. To address this issue, this paper introduces a novel visual-language collaborative representation network (MCRNet), aiming at acquiring a generalized capability for collaborative fusion and representation of multi-modal information. Specifically, MCRNet learns to generate relational matrices from an information fusion perspective to acquire aligned multi-modal features. This relationship generation strategy is category-agnostic, so it can be generalized to new domains. A class-adaptive fine-tuning inference technique is also introduced to help MCRNet efficiently learn alignment knowledge for new categories using limited data. Additionally, the paper establishes a new broad-domain few-shot image classification benchmark containing seven evaluation datasets from five domains. Comparative experiments demonstrate that MCRNet outperforms current state-of-the-art models, achieving an average improvement of 13.06% and 13.73% in the 1-shot and 5-shot settings, highlighting the superior performance and applicability of MCRNet across various domains. © 2024 ACM.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

A Deployment of Ensemble-Based IDS on Raspberry Pi 13

A Deployment of Ensemble-Based IDS on Raspberry Pi

引用

13th IEEE Global Conference on Consumer Electronic, GCCE 2024

作者： Musthafa, Muhammad Bisri Huda, Samsul Shamim, S.M. Kodera, Yuta Mwaura, Jedidah Araki, Shunsuke Nogami, Yasuyuki Okayama University Graduate School of Environmental Life Natural Science and Technology Japan Okayama University Green Innovation Center Japan Kyushu Institute of Technology Graduate School of Computer Science and Systems Engineering Fukuoka Japan Kyushu Institute of Technology Faculty of Computer Science and Systems Engineering Fukuoka Japan

ISBN: (纸本)9798350355079

Intrusion Detection Systems (IDS) are essential for safeguarding IoT networks against various attacks. Our previously developed ensemble-based IDS model, which combines stacked Long Short-Term Memory (LSTM) networks with ANOVA feature selection, achieved a remarkable accuracy of 97.46% and an overfitting value of 0.22% in classifying network attacks. However, deploying this powerful model on resource-constrained devices like Raspberry Pi poses challenges due to limited computational resources. In this paper, we tackle these challenges by converting the original model to TensorFlow Lite (TFLite) and applying quantization techniques to optimize the model size for deployment on a Raspberry Pi 3b+. Our approach significantly reduces the model size by 90.9% while maintaining an impressive accuracy of 96.96%, only a 0.5% reduction from the original model. The optimized model demonstrates efficient performance on the Raspberry Pi, with a loading time of 100.2ms and an inference time of 1.2ms. © 2024 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Research on the Transient Potential Rise Characteristics of High - speed Railway Rail Lightning Strike and Its Impact on Trackside Signal Equipment 37

Research on the Transient Potential Rise Characteristics of ...

引用

37th International Conference on Lightning Protection, ICLP 2024

作者： Wang, Shulai Xiang, Nianwen Wang, Chaoqun Lyu, Zengwei Chen, Zhiwei School of Electrical Engineering and Automation Hefei University of Technology Hefei China Institute of Energy Hefei Comprehensive National Science Center Hefei China School of Computer Science and Information Engineering Hefei University of Technology Hefei China

ISBN: (纸本)9783800764174

Lightning strikes on the catenary will cause a sharp rise in rail potential, which directly threatens the safety of the signal equipment connected to the rail along the trackside. It is urgent to carry out research on the propagation law of transient overvoltage on rails caused by lightning strikes on the catenary. A rail lightning electromagnetic transient calculation model considering the influence of integrated grounding system and catenary is established, and the influence law of different lightning strike point positions, different pier heights, and different lightning current waveforms on the distribution of rail potential rise is calculated. The specific conclusions are as follows: when lightning strikes different positions within a catenary span, the rail potential within 700m near the lightning strike point will exceed the lightning strike tolerance level of the trackside signal equipment, causing equipment damage. The potential rise increases by about 9 kV;for every 5 m increase in pier height, the rail potential rise increases by about 16%. Based on the law of the rail potential rise, the common-mode protection methods for the trackside signal equipment are proposed by appropriately reducing the horizontal connection distance of uplink and downlink ground wires to decrease the rail potential and installing common-mode protection devices. © 2024 IEEE, © VDE VERLAG GMBH ∙ Berlin ∙ Offenbach.

关键词： Signaling equipment

来源：评论

学校读者我要写书评

暂无评论

Enhanced Experts with Uncertainty-Aware Routing for Multimodal Sentiment Analysis 24

Enhanced Experts with Uncertainty-Aware Routing for Multimod...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Gao, Zixian Hu, Disen Jiang, Xun Lu, Huimin Shen, Heng Tao Xu, Xing Center for Future Media School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China School of Automation Southeast University Nanjing China College of Electronic and Information Engineering Tongji University Shanghai China

ISBN: (纸本)9798400706868

Multimodal sentiment analysis, which has garnered widespread attention in recent years, aims to predict human emotional states using multimodal data. Previous studies have primarily focused on enhancing multimodal fusion and integrating information across different modalities, while overlooking the impact of noisy data on the internal features of each single modality. In this paper, we propose the Enhanced experts with Uncertainty-Aware Routing (EUAR) method to address the influence of noisy data on multimodal sentiment analysis by capturing uncertainty and dynamically altering the network. Specifically, we introduce the Mixture of Experts approach into multimodal sentiment analysis for the first time, leveraging its properties under conditional computation to dynamically alter the network in response to different types of noisy data. Particularly, we refine the experts within the MoE framework to capture uncertainty in the data and extract clearer features. Additionally, a novel routing mechanism is introduced. Through our proposed U-loss, which utilizes the quantified uncertainty by experts, the network learns to route different samples to experts with lower uncertainty for processing, thus obtaining clearer, noise-free features. Experimental results demonstrate that our method achieves state-of-the-art performance on three widely used multimodal sentiment analysis datasets. Moreover, experiments on noisy datasets show that our approach outperforms existing methods in handling noisy data. © 2024 ACM.

关键词： Data assimilation

来源：评论

学校读者我要写书评

暂无评论

Enhancing Banking Security: An Integrated Approach to IoT Threat Mitigation with Artificial Intelligence 5th

Enhancing Banking Security: An Integrated Approach to IoT Th...

引用

5th International Conference on Recent Trends in Machine Learning, IoT, Smart Cities, and Applications, ICMISC 2024

作者： Rhaif Alsahlanee, Abbas Thajeel Singh, Jagendra Garg, Neha Rajpoot, Abha Kiran Tiwari, Mohit Elangovan, Muniyandy Development and Continuous Education Center University of Thi-Qar Thi-Qar Iraq School of Computer Science Engineering and Technology Bennett University Greater Noida India School of Computer Science Engineering & Technology Bennett University Greater Noida India School of Computing Science and Engineering Galgotias University Greater Noida India Department of Computer Science and Engineering Bharati Vidyapeeth’s College of Engineering Delhi India Department of Biosciences Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai India

ISBN: (纸本)9789819788644

This research presents a composite approach to bank security issues by integrating the Internet of Things (IoT) threat reduction and artificial intelligence (AI) into systems. In this research, machine learning (ML) techniques are deployed across key banking operations such as ATM false alarms, health monitoring of hardware components, and adding security to locker rooms. Through the utilization of ML models such as CNN, SVM, KNN, and RNN, significant advancements are achieved in detecting fraudulent activities within ATM transactions. ML-driven approaches have high precision, recall and accuracy rates in bolstering ATM network security. In addition, predictive maintenance models trained on sensor data from ATM hardware components enable proactive maintenance interventions. This in turn leads to improved ATM availability, reduced downtime, and cost savings. Biometric authentication systems based on ML models appear to have promising effects in locker room security. It is aimed at enabling security measures to be improved intelligently. Overall, this research shows how banks can use AI to tackle threats such as those maxing home machines or video-recordings by infiltrating introductory courses. By making the most of IoT-linked data and ML algorithms, the ability of banks to detect and prevent fraudulent activities can be improved in addition to having a preventive maintenance program where fees are paid in ATM withdrawals. These advancements contribute not only to safeguarding the financial assets and customer data but also to fostering trust and confidence in the banking systems. Continual research and innovation in AI and IoT technologies are crucial in order to strengthen banking security measures against the digital challenges. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Predictive maintenance

来源：评论

学校读者我要写书评

暂无评论

Identification of Ischemic Stroke Origin Using Machine Learning Techniques

Identification of Ischemic Stroke Origin Using Machine Learn...

引用

International Conference on Advances in Data-driven Computing and Intelligent Systems, ADCIS 2022

作者： Prasad, Rajan Shukla, Praveen Kumar Artificial Intelligence Research Center Department of Computer Science and Engineering School of Engineering Babu Banarasi Das University Lucknow India

ISBN: (纸本)9789819909803

An ischemic stroke arises as a result of atherosclerosis. It promotes tissue and cholesterol accumulation in the vascular system. When a large amount of plaque collects in one region, it has the potential to reduce blood supply to crucial organs. Carotid arteries are blood vessels that carry blood to the brain and are found in the neck. A stroke may develop if plaque plugs a carotid artery (carotid artery disease). Atherosclerosis is mostly asymptomatic. As a result, many people are unaware they have atherosclerosis until they have a stroke or suffer the additional repercussions of clogged arteries near other organs, such as a heart attack. The primary goal of this project is to identify cases of ischemic stroke by utilizing blood clot imaging datasets with CNN and CatBoost classifiers. The experimental results show that the CNN achieved 97.5% accuracy, whereas the CatBoost achieved 73.5% accuracy. This study will aid healthcare providers in properly detecting the sources of blood clots in fatal strokes, allowing doctors to deliver the best post-stroke treatment and minimize the risk of a second stroke. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Classification (of information)

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：