检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Ye, Sheng Hu, Yubin Lin, Matthieu Wen, Yu-Hui Zhao, Wang Liu, Yong-Jin Wang, Wenping BNRist The Department of Computer Science and Technology Tsinghua University China Beijing Key Laboratory of Traffic Data Analysis and Mining School of Computer and Information Technology Beijing Jiaotong University China The Department of Computer Science & Engineering Texas A&M University United States

The reconstruction of indoor scenes from multi-view RGB images is challenging due to the coexistence of flat and textureless regions alongside delicate and fine-grained regions. Recent methods leverage neural radiance fields aided by predicted surface normal priors to recover the scene geometry. These methods excel in producing complete and smooth results for floor and wall areas. However, they struggle to capture complex surfaces with high-frequency structures due to the inadequate neural representation and the inaccurately predicted normal priors. This work aims to reconstruct high-fidelity surfaces with fine-grained details by addressing the above limitations. To improve the capacity of the implicit representation, we propose a hybrid architecture to represent low-frequency and high-frequency regions separately. To enhance the normal priors, we introduce a simple yet effective image sharpening and denoising technique, coupled with a network that estimates the pixel-wise uncertainty of the predicted surface normal vectors. Identifying such uncertainty can prevent our model from being misled by unreliable surface normal supervisions that hinder the accurate reconstruction of intricate geometries. Experiments on the benchmark datasets show that our method outperforms existing methods in terms of reconstruction quality. Furthermore, the proposed method also generalizes well to real-world indoor scenarios captured by our hand-held mobile phones. Our code is publicly available at: https://***/yec22/Fine-Grained-Indoor-Recon. Copyright © 2023, The Authors. All rights reserved.

关键词： Surface reconstruction

来源：评论

学校读者我要写书评

暂无评论

CONCRETIZER: MODEL INVERSION ATTACK VIA OCCUPANCY CLASSIFICATION AND DISPERSION CONTROL FOR 3D POINT CLOUD RESTORATION

arXiv

引用

arXiv 2025年

作者： Kim, Youngseok Hwang, Sunwook Kim, Hyung-Sin Bahk, Saewoong Department of Electrical and Computer Engineering Seoul National University Korea Republic of System LSI Samsung Electronics Korea Republic of Graduate School of Data Science Seoul National University Korea Republic of

The growing use of 3D point cloud data in autonomous vehicles (AVs) has raised serious privacy concerns, particularly due to the sensitive information that can be extracted from 3D data. While model inversion attacks have been widely studied in the context of 2D data, their application to 3D point clouds remains largely unexplored. To fill this gap, we present the first in-depth study of model inversion attacks aimed at restoring 3D point cloud scenes. Our analysis reveals the unique challenges, the inherent sparsity of 3D point clouds and the ambiguity between empty and non-empty voxels after voxelization, which are further exacerbated by the dispersion of non-empty voxels across feature extractor layers. To address these challenges, we introduce ConcreTizer, a simple yet effective model inversion attack designed specifically for voxel-based 3D point cloud data. ConcreTizer incorporates Voxel Occupancy Classification to distinguish between empty and non-empty voxels and Dispersion-Controlled Supervision to mitigate non-empty voxel dispersion. Extensive experiments on widely used 3D feature extractors and benchmark datasets, such as KITTI and Waymo, demonstrate that ConcreTizer concretely restores the original 3D point cloud scene from disrupted 3D feature data. Our findings highlight both the vulnerability of 3D data to inversion attacks and the urgent need for robust defense strategies. Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Route-and-Aggregate Decentralized Federated Learning Under Communication Errors

arXiv

引用

arXiv 2025年

作者： Li, Weicai Lv, Tiejun Ni, Wei Zhao, Jingbo Hossain, Ekram Poor, H. Vincent Beijing100876 China School of Electrical and Data Engineering University of Technology SydneyNSW Australia School of Electrical and Data Engineering University of Technology SydneyNSW2007 Australia The School of Computing Science and Engineering University of New South Wales NSW2052 Australia Department of Electrical and Computer Engineering University of Manitoba Canada Department of Electrical and Computer Engineering Princeton University PrincetonNJ08544 United States

Decentralized federated learning (D-FL) allows clients to aggregate learning models locally, offering flexibility and scalability. Existing D-FL methods use gossip protocols, which are inefficient when not all nodes in the network are DFL clients. This paper puts forth a new D-FL strategy, termed Route-and-Aggregate (R&A) D-FL, where participating clients exchange models with their peers through established routes (as opposed to flooding) and adaptively normalize their aggregation coefficients to compensate for communication errors. The impact of routing and imperfect links on the convergence of R&A D-FL is analyzed, revealing that convergence is minimized when routes with the minimum end-to-end packet error rates are employed to deliver models. Our analysis is experimentally validated through three image classification tasks and two next-word prediction tasks, utilizing widely recognized datasets and models. R&A DFL outperforms the flooding-based D-FL method in terms of training accuracy by 35% in our tested 10-client network, and shows strong synergy between D-FL and networking. In another test with 10 D-FL clients, the training accuracy of R&A D-FL with communication errors approaches that of the ideal C-FL without communication errors, as the number of routing nodes (i.e., nodes that do not participate in the training of D-FL) rises to 28. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Customised Neural Network Model and ML Classifiers for Brain Stroke Prediction

Comparative Analysis of Customised Neural Network Model and ...

引用

Network, Multimedia and Information Technology (NMITCON) International Conference on

作者： Tejashree S Varshitha S Ashwini Kodipalli Trupthi Rao Kumarswamy S Department of CSE Global Academy of Technology Bangalore Karnataka India Artificial Intelligence and Data Science and Engineering Global Academy of Technology Bangalore Karnataka India Department of Computer Science and Engineering Global Academy of Technology Bangalore Karnataka India

An abrupt interruption of the brain's blood supply leads to stroke, which causes severe damage to the brain. Due to this condition, there are people disabled and thus leading to death. In this work, the dataset contains ten features where environmental tests are done for the stroke which is utilized to evaluate the efficiency of Machine Learning classifiers and Deep Learning Artificial Neural Network. The purpose of this paper is to provide a forecast of the likelihood of a brain stroke occurring at an early stage using Machine Learning and Deep Learning techniques. The accuracy and efficiency of various types of Machine Learning models and customized Deep Learning models have been compared based on this impression. The custom Deep Learning model here used is an Artificial Neural Network with dropouts having the highest accuracy of 97.06% using RMSprop optimizer and a Decision Tree Machine Learning Classifier, achieved the highest accuracy of 95.98%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Contactless attendance management system using artificial intelligence 2

Contactless attendance management system using artificial in...

引用

2nd International Conference on Smart and Intelligent Learning for Information Optimization, CONSILIO 2020

作者： Rajamanogaran, M. Subha, S. Baghavathi Priya, S. Sivasamy, Jeevitha Data Science Consultant Chennai Tamil Nadu India Department of Computer Science and Engineering Rajalakshmi Institute of Technology Chennai Tamil Nadu India Department of Information Technology Rajalakshmi Engineering College Chennai Tamil Nadu India Department of Computer and Information Science Annamalai University Chidambaram Tamil Nadu India

Attendance monitoring is one of the vital administrative processes in all educational institutions and organizations. A well-structured system will enable the institutions to have an increasing growth. It helps the students and teachers in all ways to have a good progress in the attendance, thereby reducing the teachers’ time and effort. In physical classrooms, traditional method of calling the students names and marking their presence/absence is the routine process followed regularly. Attendance monitoring system is proposed using artificial intelligence. Firstly, a database containing the facial images of the students in a particular class is constructed. Knowledge gained using Convolutional Neural Network (CNN) is reused in a perfect manner using transfer learning. This system is designed to improve the students’ engagement time inside the classroom, to communicate to the parents frequently, to avoid proxy attendance and to generate detailed reports for future reference. © 2021 Institute of Physics Publishing. All rights reserved.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Implementing and Analyzing Behaviour of ML based Tamil Songs Recommendation System 2

Implementing and Analyzing Behaviour of ML based Tamil Songs...

引用

2nd International Conference on Sustainable Computing and data Communication Systems, ICSCDS 2023

作者： Garg, Shalini Pooja Chaoudhary, Shailee Lohmor Dixit, Rinku Govil, Nikhil Design and Technology University Mit Art MH Pune India Galgotias University School of Cse Delhi - NCR Gr. Noida India New Delhi Institute of Management Department of Aiml - Data Science New Delhi India Gla University Department of Computer Engineering & Applications U.P. Mathura India

ISBN: (纸本)9781665491990

Music gives listeners a chance to relax, relieve fatigue and connect one culture with another. In such a situation, if the favorite song-music is heard one after another and according to the emotion, it gives a unique and wonderful experience to the listener. It is determined based on the emotion and interest, ratings provided by listeners and search patterns. In this research article, the proposed method presents Tamil songs compilation and recommendation system with the help of these factors. The behavior of this recommendation system is also analyzed in this system. The proposed system is developed in Java programming language by adopting the techniques of Machine Learning. Through this system, listeners have the choice to play category wise songs. As per the listening history and patterns, upcoming songs are predicted to the listeners. This system also provides ratings to each song. This developed application can also be used commercially as it has been validated that the ratings provided by the listeners are in line with IMDb ratings. © 2023 IEEE.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Self-supervised Topic Taxonomy Discovery in the Box Embedding Space

arXiv

引用

arXiv 2024年

作者： Lu, Yuyin Chen, Hegang Mao, Pengbo Rao, Yanghui Xie, Haoran Wang, Fu Lee Li, Qing School of Computer Science and Engineering Sun Yat-sen University Guangzhou China School of Data Science Lingnan University Hong Kong School of Science and Technology Hong Kong Metropolitan University Hong Kong Department of Computing The Hong Kong Polytechnic University Hong Kong

Topic taxonomy discovery aims at uncovering topics of different abstraction levels and constructing hierarchical relations between them. Unfortunately, most of prior work can hardly model semantic scopes of words and topics by holding the Euclidean embedding space assumption. What’s worse, they infer asymmetric hierarchical relations by symmetric distances between topic embeddings. As a result, existing methods suffer from problems of low-quality topics at high abstraction levels and inaccurate hierarchical relations. To alleviate these problems, this paper develops a Box embedding-based Topic Model (BoxTM) that maps words and topics into the box embedding space, where the asymmetric metric is defined to properly infer hierarchical relations among topics. Additionally, our BoxTM explicitly infers upper-level topics based on correlation between specific topics through recursive clustering on topic boxes. Finally, extensive experiments validate high-quality of the topic taxonomy learned by BoxTM. Copyright © 2024, The Authors. All rights reserved.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

UpCycling: Semi-supervised 3D Object Detection without Sharing Raw-level Unlabeled Scenes

UpCycling: Semi-supervised 3D Object Detection without Shari...

引用

International Conference on computer Vision (ICCV)

作者： Sunwook Hwang Youngseok Kim Seongwon Kim Saewoong Bahk Hyung-Sin Kim Department of Electrical and Computer Engineering Seoul National University SK Telecom Seoul Korea Graduate School of Data Science Seoul National University

Semi-supervised Learning (SSL) has received increasing attention in autonomous driving to reduce the enormous burden of 3D annotation. In this paper, we propose UpCycling, a novel SSL framework for 3D object detection with zero additional raw-level point cloud: learning from unlabeled de-identified intermediate features (i.e., "smashed" data) to preserve privacy. Since these intermediate features are naturally produced by the inference pipeline, no additional computation is required on autonomous vehicles. However, generating effective consistency loss for unlabeled feature-level scene turns out to be a critical challenge. The latest SSL frameworks for 3D object detection that enforce consistency regularization between different augmentations of an unlabeled raw-point scene become detrimental when applied to intermediate features. To solve the problem, we introduce a novel combination of hybrid pseudo labels and feature-level Ground Truth sampling (F-GT), which safely augments unlabeled multi-type 3D scene features and provides high-quality supervision. We implement UpCycling on two representative 3D object detection models: SECOND-IoU and PV-RCNN. Experiments on widely-used datasets (Waymo, KITTI, and Lyft) verify that UpCycling outperforms other augmentation methods applied at the feature level. In addition, while preserving privacy, UpCycling performs better or comparably to the state-of-the-art methods that utilize raw-level unlabeled data in both domain adaptation and partial-label scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Controller Optimization of an Automated Force Calibration Device Based on an Improved Particle Swarm Algorithm

Controller Optimization of an Automated Force Calibration De...

引用

Sensors and Information Technology (ICSI), International Conference on

作者： Jiming Jiang Xiang Dong Qiang Li Huibin Cao School of Electrical Engineering and Automation Anhui University Hefei China Department of Computer Science and Technology College of Big Data and Internet Shenzhen Technology University Shenzhen China Hefei Institutes of Physical Science Chinese Academy of Sciences Hefei Anhui China

ISBN: (数字)9798331542887

ISBN: (纸本)9798331542894

This paper presents a PID controller optimization method based on an Improved Particle Swarm Optimization (IPSO) algorithm, aimed at enhancing the design and performance of automated force calibration devices. Traditional force calibration systems often suffer from limitations in automation and force application accuracy, which this method seeks to address. The IPSO algorithm incorporates logistic chaotic mapping to initialize particle positions and enhances the standard particle swarm algorithm through nonlinear decreasing weights and adaptive learning factors, resulting in improved global search capabilities and faster local convergence. Additionally, the objective function is refined by normalizing and weighting the integral of time-multiplied absolute error (ITAE) and the overshoot, thereby further enhancing the force application accuracy of the PID controller. Simulation results demonstrate that the IPSO algorithm offers significant advantages in terms of settling time, overshoot, and ITAE error when compared to traditional PID tuning methods (e.g., Ziegler-Nichols) and other optimization algorithms (e.g., GA, GWO). Experimental results validate the effectiveness and practicality of the IPSO algorithm, providing an efficient solution for the automation and high-precision control of force sensor calibration systems.

关键词： Performance evaluation Accuracy Automation Force Optimization methods Systems modeling Calibration Particle swarm optimization Tuning Standards

来源：评论

学校读者我要写书评

暂无评论

Hybrid Movie Recommendation System using Neural Collaborative Filtering with Gradient Decent and Association Rule Mining

Hybrid Movie Recommendation System using Neural Collaborativ...

引用

Software, Systems and Information Technology (SSITCON), International Conference on

作者： J Cindhamani Rajkumar Pogaku M. P Sahana Myasar M. Adnan M. Suganya Department of Artificial Intelligence and Data Science Vel Tech Multi Tech Dr. Rangrajan Dr. Sakunthala Engineering College Avadi Chennai India School of Computer Science and Engineering Malla Reddy Engineering College for Women (UGC-Autonomous) Maisammaguda Hyderabad India Department of Computer Science and Engineering Dayananda Sagar College of Engineering Bengaluru India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of Electronics and Communication Engineering New Prince Shri Bhavani College of Engineering and Technology Chennai India

ISBN: (数字)9798350352931

ISBN: (纸本)9798350352948

In recent days, most of the people are entertained by watching movies and shows, so developing a recommendation system help people across the world in saving the time to search for high rated movies. Usually, movies are recommended based on the user interests in the different movie genres such as horror, drama, romantic and comedy. Previous researchers suggested various methods for movie recommendation using Content Based Filtering (CBF), Collaborative Filtering (CF) which failed to predict the non-linear relationship recommendation due to the availability of many genres. To overcome this challenge, we proposed hybrid model namely Neural Collaborative Filtering (NCF) with gradient decent and Association Rule Mining (ARM). Initially, MovieLens dataset is employed and then data cleaning data organization is incorporated in pre-processing. Further, the features are extracted through the Term Frequency-Inverse Document Frequency (TF-IDF) and Latent Semantic Analysis (LSA) to convert the text data into numerical data and increase the performance respectively. Finally, the developed model was measured using Root mean square error (RMSE), precision, recall, and F 1 score. Hence, the proposed hybrid recommendation model provides higher accuracy value when compared to Knowledge-aware Attentional Neural Network (KANN).

关键词： Collaborative filtering Semantics Neural networks Organizations Motion pictures Feature extraction Frequency conversion Numerical models Association rule learning Recommender systems

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：