检索结果-内蒙古大学图书馆

Diff-Reg v2: Diffusion-Based Matching Matrix Estimation for Image Matching and 3D Registration

学校读者我要写书评

暂无评论

arXiv

arXiv 2025年

作者： Wu, Qianliang Jiang, Haobo Ding, Yaqing Luo, Lei Xie, Jin Yang, Jian PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education China School of Computer Science and Engineering Nanjing University of Science and Technology China Nanyang Technological University Singapore Visual Recognition Group Faculty of Electrical Engineering Czech Technical University in Prague Prague Czech Republic State Key Laboratory for Novel Software Technology Nanjing University Nanjing China School of Intelligence Science and Technology Nanjing University Suzhou China

Establishing reliable correspondences is crucial for all registration tasks, including 2D image registration, 3D point cloud registration, and 2D-3D image-to-point cloud registration. However, these tasks are often complicated by challenges such as scale inconsistencies, symmetry, and large deformations, which can lead to ambiguous matches. Previous feature-based and correspondence-based methods typically rely on geometric or semantic features to generate or polish initial potential correspondences. Some methods typically leverage specific geometric priors, such as topological preservation, to devise diverse and innovative strategies tailored to a given enhancement goal, which cannot be exhaustively enumerated. Additionally, many previous approaches rely on a single-step prediction head, which can struggle with local minima in complex matching scenarios. To address these challenges, we introduce an innovative paradigm that leverages a diffusion model in matrix space for robust matching matrix estimation. Our model treats correspondence estimation as a denoising diffusion process in the matching matrix space, gradually refining the intermediate matching matrix to the optimal one. Specifically, we apply the diffusion model in the doubly stochastic matrix space for 3D-3D and 2D-3D registration tasks. In the 2D image registration task, we deploy the diffusion model in a matrix sub-space, where dual-softmax projection regularization is applied. For all three registration tasks, we provide adaptive matching matrix embedding implementations tailored to the specific characteristics of each task while maintaining a consistent"match-to-warp" encoding pattern. Furthermore, we adopt a lightweight design for the denoising module. In inference, once points or image features are extracted and fixed, this module performs multi-step denoising predictions through reverse sampling. Evaluations on both 2D and 3D registration tasks demonstrate the effectiveness of our approach. Copyrigh

关键词： Stochastic systems

A novel Offloading Approach of Iot User Perception Task Based on Quantum Behavior Particle Swarm Optimization

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Chen, Lu Zhang, De-Gan Zhang, Jie Zhang, Ting Wang, Wen-Jing Cao, Ya-Hui Tianjin Key Lab of Intelligent Computing and Novel Software Technology Tianjin University of Technology Tianjin300384 China School of Electronic and Information Engineering Beijing Jiaotong University Beijing100044 China School of Sports Economics and Management Tianjin University of Sport Tianjin301617 China

To solve the problem of high time consumption and energy loss of edge computing perceptual task offloading in the mobile IoT (Internet of things), it is necessary to increase the particle population diversity and the performance of escaping the local optimum solution in the intelligent optimization algorithm. For the offloading scenario of multi edge central servers and smart devices of multi IoT users, the NQPSOOM method (A new method of user perceptual task offloading in the IoT based on optimizing strategy of quantum behavior particle population) was proposed. The adaptability variance of the quantum behavior particle population was designed to measure the convergence of the particle population. The logistic chaos perturbation strategy was used to improve the performance of the intelligent optimization algorithm of escaping the local optimum solution. The exchange and mutation operations of fast elite non-dominated sorting genetic algorithm based on crowding coefficient were designed to select the particle population with a high diversity of quantum behavior particle population. Compared with the other five perceptual task computing offloading methods in recent three years, the NQPSOOM method reduces the adaptability of the intelligent optimization algorithm, decreases the time consumption and energy loss of the algorithm. © 2022, The Authors. All rights reserved.

关键词： Internet of things

New Method of Vehicle Cooperative Communication Based on Fuzzy Logic and Signal Game Strategy

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Zhang, De-Gan Ni, Chen-Hao Zhang, Jie Zhang, Ting Zhang, Zhi-Hao Su, Hao Tianjin Key Lab of Intelligent Computing and Novel Software Technology Tianjin University of Technology Tianjin300384 China School of Electronic and Information Engineering Beijing Jiaotong University Beijing100044 China School of Sports Economics and Management Tianjin University of Sport Tianjin301617 China

With the rapid development of the Internet of Vehicles, at the same time, due to the limited storage space of vehicles and more and more data requirements, vehicles need more and more vehicle-to-vehicle collaboration. At this time, a vehicular ad-hoc network (VANET) is required for inter-vehicle communication cooperation, and link breaks often occur due to the rapid movement of vehicles and the uncertainty of vehicle communication quality, which affects the communication quality. Because of these situations, this paper proposes a vehicle cooperative communication method based on fuzzy logic and signal game strategy, obtains more real vehicle position and speed information on the Kalman filter, and then selects relay nodes to assist in requesting information message distribution of fuzzy logic. Each time the vehicle receives a request, the first data information is returned to the signal game. If all the information requests are not received the first time, the information requesting vehicle will send the second information request according to the signal strength and the beliefs observed on the previous vehicle. The experimental verification shows that this method can realize the function of vehicle cooperative communication, solve the problems of low VANET link stability and broadcast storm, and has a good supporting role in establishing the practicality of vehicle cooperation in VANET. © 2022, The Authors. All rights reserved.

关键词： Fuzzy logic

Combining Bilstm and Self-Attention Mechanism for Code Search

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Cao, Ben Liu, Jianxun School of Computer Science and Technology Hunan University of Science and Technology Hunan411100 China Hunan Provincial Key Lab. for Services Computing and Novel Software Technology Hunan University of Science and Technology Hunan411100 China

With the wide application of deep learning in code search, especially the proposed code search model based on attention mechanism, the accuracy of code search has been greatly improved. However, the attention mechanism only captures the attention weight relationship between two words in the code fragment, without considering the contextual semantic relationship that exists between words in the code fragment, which can help improve the accuracy of code search. To address this problem, this paper proposes a model that combining BiLSTM and self-attention mechanisms for code search(CBLSAM-CS). The model first captures the contextual relationship of each word in the code fragment by a recurrent neural network, and then uses the self-attentive mechanism to extract deep-level features of the sequence. In order to verify the effectiveness of the proposed model, the paper conducts an experimental comparison with three other baseline models, on the basis of a public dataset containing 18 million code *** experimental results show that the proposed model in this paper achieves 92.4% and 93.5% in MRR value and NDCG metrics, respectively, which are better than the baseline ***, it shows that the CBLSAM-CS model proposed in this paper can effectively improve the accuracy and efficiency of code search. © 2022, The Authors. All rights reserved.

关键词： Recurrent neural networks

Multi-Layer Self-Attention Residual Network for Code Search

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Hu, Haize Professor, Jianxun Liu School of Computer Science and Technology Hunan University of Science and Technology Hunan411100 China Hunan Provincial Key Lab. for Services Computing and Novel Software Technology Hunan University of Science and Technology Hunan411100 China

To improve the efficiency of programming, software developers usually search existing code snippets in open source code repositories to modify and reuse them. Therefore, how to get the right code snippet from the open source code repository quickly and accurately is the focus of current software development research. Nowadays, code search is one of the solutions. Traditional information retrieval techniques have been gradually abandoned by researchers due to their low search efficiency and accuracy. In recent years, the code search methods based on machine learning alternatives have been widely used and researched due to the better search results. Various code search models have been proposed, especially the self-attention model, which has greatly improved the accuracy of code search. The self-attentive model for feature extraction mainly relies on the weights of each word in the code sequence unit, while ignoring the structural relationships among the code sequences, which contain rich feature information in these components. To compensate the shortage of self- attention mechanism, a multi-layer self- attention residual network based code search model (MSARN-CS) is proposed in this paper. In the MSARN-CS model, not only the weight of each word in the code sequence unit is considered, but also the effect of embedding between code sequence units is calculated. In addition, an optimization model of residuals is introduced to compensate for the loss of information in the code sequences during the model training. To verify the search effectiveness of the MSARN-CS model, three other baseline models are compared on the basis of extensive source code data. The experimental results show that the MSARN-CS model has better search results compared with the baseline model. Therefore, the MSARN-CS model proposed in the paper can effectively improve the accuracy of code search and enhance the programming efficiency of developers. © 2022, The Authors. All rights reserved.

关键词： Efficiency

Modeling Multilevel Business Process Monitoring via BPMN Extension

学校读者我要写书评

暂无评论

Concurrency and Computation: Practice and Experience 2025年第12-14期37卷

作者： Kang, Guosheng Wang, Zhen Cheng, Hangyu Liu, Jianxun Wen, Yiping Peng, Jun School of Computer Science and Engineering Hunan University of Science and Technology Xiangtan China Hunan Provincial Key Lab. for Services Computing and Novel Software Technology Hunan University of Science and Technology Xiangtan China School of Computer Science and Technology Xi'an Jiaotong University Xi'an China School of Computer Science Fudan University Shanghai China

Business process monitoring involves real-time supervision of a series of activities carried out by an organization to achieve specific objectives. Process event monitoring points (PEMP) are used to pinpoint specific locations within a process model where expected events are anticipated to occur. However, in general business process modeling languages, such as business process model and notation (BPMN), there is a lack of explicit modeling for PEMPs. This article proposes a method for modeling pairwise event monitoring points in business process models to track the execution status of specific activities or process segments via BPMN extension. Specifically, process monitoring points are designed by expanding the modeling element of sequence flow. Moreover, the business process model is decomposed using the refined process structure tree (RPST) to verify the soundness of the designed pairwise monitoring points. The proposed method allows flexible monitoring at the process segment level. And the process monitoring data could be used for a clear understanding of business progress, especially useful in heterogeneous business process model to support decision-making. Through case study from real-world business processes, the effectiveness and usefulness of the proposed process monitoring modeling method is validated. © 2025 John Wiley & Sons Ltd.

关键词： Electronic trading

Disentangled Hypergraph Collaborative Filtering for Social Recommendation

学校读者我要写书评

暂无评论

Disentangled Hypergraph Collaborative Filtering for Social R...

IEEE International Conference on Web Services (ICWS)

作者： Xiao Liu Shunmei Meng Qianmu Li Xiaolong Xu Lianyong Qi Wanchun Dou Jing Zhang Xuyun Zhang Nanjing University of Science and Technology China State Key Lab. for Novel Software Technology Nanjing University China School of Computer and Software Nanjing University of Information Science and Technology China College of Computer Science and Technology China University of Petroleum (East China) China School of Cyber Science and Engineering Southeast University China School of Computing Macquarie University Australia

In the current era of information overload, service recommendations have emerged as a valuable tool for enhancing the user experience. Among them, social recommendation models have shown promising results by incorporating social relationships to improve representation learning. However, most of these models lack fine-grained modeling of social user behavior, leading to a unified representation of users and a loss of expressiveness in user representations. To address this issue, we propose DisenHGCF, a new social recommendation approach based on disentangled hypergraph collaborative filtering, to disentangle the representations of users and items at the granularity of social users’ intents. This approach aims to provide a more nuanced understanding of the user, leading to more accurate and personalized recommendations. To be specific, DisenHGCF leverages hypergraphs to represent the complex relationships among users, friends, and items. By using the hypergraph disentangling module based on attention, it is able to disentangle the user’s intents and generate users representations in cooperating their intents for recommendation tasks. Additionally, a contrastive learning task based on intent-weight perturbation is designed to enhance representational learning. The experimental results obtained from the BeiBei and Beidian datasets demonstrate the superiority of our proposed approach in comparison to previous baseline methods, as evidenced by higher Recall and NDCG scores.

关键词：

JobFormer: Skill-Aware Job Recommendation with Semantic-Enhanced Transformer

学校读者我要写书评

暂无评论

arXiv

arXiv 2024年

作者： Guan, ZhiHao Yang, Jia-Qi Yang, Yang Zhu, Hengshu Li, Wenjie Xiong, Hui the Nanjing University of Science and Technology Nanjing210094 China State Key Laboratory for Novel Software Technology Nanjing University Nanjing210023 China Career Science Lab BOSS Zhipin Beijing100028 China the Hong Kong Polytechnic University Hong Kong the Artificial Intelligence Thrust The Hong Kong University of Science and Technology Guangzhou China PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology China

Job recommendation aims to provide potential talents with suitable job descriptions (JDs) consistent with their career trajectory, which plays an essential role in proactive talent recruitment. In real-world management scenarios, the available JD-user records always consist of JDs, user profiles, and click data, in which the user profiles are typically summarized as the user’s skill distribution for privacy reasons. Although existing sophisticated recommendation methods can be directly employed, effective recommendation still has challenges considering the information deficit of JD itself and the natural heterogeneous gap between JD and user profile. To address these challenges, we proposed a novel skill-aware recommendation model based on the designed semantic-enhanced transformer to parse JDs and complete personalized job recommendation. Specifically, we first model the relative items of each JD and then adopt an encoder with the local-global attention mechanism to better mine the intra-job and inter-job dependencies from JD tuples. Moreover, we adopt a two-stage learning strategy for skill-aware recommendation, in which we utilize the skill distribution to guide JD representation learning in the recall stage, and then combine the user profiles for final prediction in the ranking stage. Consequently, we can embed rich contextual semantic representations for learning JDs, while skill-aware recommendation provides effective JD-user joint representation for click-through rate (CTR) prediction. To validate the superior performance of our method for job recommendation, we present a thorough empirical analysis of large-scale real-world and public datasets to demonstrate its effectiveness and interpretability. Copyright © 2024, The Authors. All rights reserved.

关键词： User profile

Bridge the Gap between Supervised and Unsupervised Learning for Fine-Grained Classification

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Wang, Jiabao Li, Yang Wei, Xiu-Shen Li, Hang Miao, Zhuang Zhang, Rui Army Engineering University of PLA Jiangsu Nanjing210007 China PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology Jiangsu Nanjing210094 China State Key Lab. for Novel Software Technology Nanjing University Jiangsu Nanjing210023 China

Unsupervised learning technology has caught up with or even surpassed supervised learning technology in general object classification (GOC) and person re-identification (re-ID). However, it is found that the unsupervised learning of fine-grained visual classification (FGVC) is more challenging than GOC and person re-ID. In order to bridge the gap between unsupervised and supervised learning for FGVC, we investigate the essential factors for the performance gap. Furthermore, we propose a simple, effective, and practical method, termed as UFCL, to alleviate the gap. Three key issues are concerned and improved: First, we introduce a robust and powerful backbone, ResNet50-IBN, which has an ability of domain adaptation when we transfer ImageNet pre-trained models to FGVC tasks. Next, we propose to introduce HDBSCAN instead of DBSCAN to do clustering, which can generate better clusters for adjacent categories with fewer hyper-parameters. Finally, we propose a weighted feature agent and its updating mechanism to do contrastive learning, which can improve the optimization process of learning the parameters of the network. The effectiveness of our UFCL is verified on four public FGVC datasets. Under the unsupervised FGVC setting, we achieve state-of-the-art results, and analyze the key factors and the important parameters to provide a practical guidance. © 2023, The Authors. All rights reserved.

关键词： Image classification