检索结果-内蒙古大学图书馆

Addr-Net: Attention Mechanism-Based Dual-Stream Deformable Medical Image Registration Network

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Fan, Chao Zhu, Xinru Peng, Bincheng Xuan, Zhihui Zhu, Zhentong School of Artificial Intelligence and Big Data Henan University of Technology Henan Province Zhengzhou City China Key Laboratory of Grain Information Processing and Control Ministry of Education Henan Province Zhengzhou City China School of Information Science and Engineering Henan University of Technology Henan Province Zhengzhou City450001 China

Medical image registration is crucial for tumor growth monitoring, radiation therapy, and disease diagnosis. Recently, U-Net type networks have become widely used in unsupervised image registration to predict dense deformation fields. However, for image data with large deformations, these networks are prone to interference from outliers and task-irrelevant regions, leading to suboptimal registration accuracy and difficulty in maintaining the continuity of the deformation field. To address this issue, we propose a novel model called ADDR-Net. ADDR-Net leverages dual encoders and incorporates a Cross-Perceptual Attention Module (CPA) alongside a Multi-dimensional Attention Module (MDA) to enhance registration. Unlike traditional methods that concatenate fixed and moving images into a single encoder, ADDR-Net uses dual encoders to separately extract features from each image, facilitating explicit feature matching between image pairs. The CPA module, embedded within the encoder, enhances the extraction of global features over long distances. Meanwhile, the MDA module, embedded within the skip connection, filters redundant features and extracts local refined features, thereby improving the U-shaped network architecture. Experimental results demonstrate that ADDR-Net outperforms state-of-the-art methods. Specifically, compared to the Transformer-based TransMorph method, ADDR-Net increases the Dice scores on two publicly available brain datasets, IXI and OASIS, by 1.4% and 3.5%, respectively. © 2024, The Authors. All rights reserved.

关键词： Network architecture

A Multi-Stage Goal-Driven Network for Pedestrian Trajectory Prediction

学校读者我要写书评

暂无评论

A Multi-Stage Goal-Driven Network for Pedestrian Trajectory ...

Computer Vision, Image and Deep Learning (CVIDL), International Conference on

作者： Xiuen Wu Tao Wang Yuanzheng Cai Lingyu Liang George Papageorgiou Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Minjiang University Fuzhou China College of Computer and Data Science Fuzhou University Fuzhou China School of Electronic and Information Engineering South China University of Technology Guangzhou China Department of Computer Science and Engineering European University Cyprus Nicosia Cyprus

ISBN: (数字)9798350373820

ISBN: (纸本)9798350373837

Pedestrian trajectory prediction plays a pivotal role in ensuring the safety and efficiency of various applications, including autonomous vehicles and traffic management systems. This paper proposes a novel method for pedestrian trajectory prediction, called multi-stage goal-driven network (MGNet). Diverging from prior approaches relying on stepwise recursive prediction and the singular forecasting of a long-term goal, MGNet directs trajectory generation by forecasting intermediate stage goals, thereby reducing prediction errors. The network comprises three main components: a conditional variational autoencoder (CVAE), an attention module, and a multi-stage goal evaluator. Trajectories are encoded using conditional variational autoencoders to acquire knowledge about the approximate distribution of pedestrians’ future trajectories, and combined with an attention mechanism to capture the temporal dependency between trajectory sequences. The pivotal module is the multi-stage goal evaluator, which utilizes the encoded feature vectors to predict intermediate goals, effectively minimizing cumulative errors in the recursive inference process. The effectiveness of MGNet is demonstrated through comprehensive experiments on the JAAD and PIE datasets. Comparative evaluations against state-of-the-art algorithms reveal significant performance improvements achieved by our proposed method.

关键词： Deep learning Pedestrians Computational modeling Predictive models Prediction algorithms Vectors Inference algorithms

Joint Association, Beamforming, and Resource Allocation for Multi-IRS Enabled MU-MISO systems With RSMA

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Chunjie Zhang, Xuhui Xing, Huijun Xue, Liang Wang, Shuqiang Shen, Yanyan Yang, Bo Guan, Xinping Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China School of Information and Electrical Engineering Hebei University of Engineering Handan056038 China Hebei Key Laboratory of Security and Protection Information Sensing and Processing Handan056038 China Shenzhen Future Network of Intelligence Institute The School of Science and Engineering The Guangdong Provincial Key Laboratory of Future Networks of Intelligence The Chinese University of Hong Kong Guangdong Shenzhen518172 China Department of Electrical and Electronic Engineering Imperial College London LondonSW7 2AZ United Kingdom Shenzhen Future Network of Intelligence Institute The Guangdong Provincial Key Laboratory of Future Networks of Intelligence The Chinese University of Hong Kong Guangdong Shenzhen518172 China Department of Automation The Key Laboratory of System Control and Information Processing Ministry of Education Shanghai Jiao Tong University Shanghai200240 China

Intelligent reflecting surface (IRS) and rate-splitting multiple access (RSMA) technologies are at the forefront of enhancing spectrum and energy efficiency in the next generation multi-antenna communication systems. This paper explores a RSMA system with multiple IRSs, and proposes two purpose-driven scheduling schemes, i.e., the exhaustive IRS-aided (EIA) and opportunistic IRS-aided (OIA) schemes. The aim is to optimize the system weighted energy efficiency (EE) under the above two schemes, respectively. Specifically, the Dinkelbach, branch and bound, successive convex approximation, and the semidefinite relaxation methods are exploited within the alternating optimization framework to obtain effective solutions to the considered problems. The numerical findings indicate that the EIA scheme exhibits better performance compared to the OIA scheme in diverse scenarios when considering the weighted EE, and the proposed algorithm demonstrates superior performance in comparison to the baseline algorithms. Copyright © 2024, The Authors. All rights reserved.

关键词： Beamforming

Predictive auto-encoding network for blind stereoscopic image quality assessment

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Xu, Jiahua Zhou, Wei Chen, Zhibo Ling, Suiyi Le Callet, Patrick CAS Key Laboratory of Technology in Geo-Spatial Information Processing and Application System University of Science and Technology of China Hefei230027 China quipe Image Perception et Interaction Laboratoire des Sciences du Numrique de Nantes Universit de Nantes Nantes44300 France

Stereoscopic image quality assessment (SIQA) has encountered non-trivial challenges due to the fast proliferation of 3D contents. In the past years, deep learning oriented SIQA methods have emerged and achieved spectacular performance compared to conventional algorithms which are only relied on extracting hand-crafted features. However, most existing deep SIQA evaluators are not specifically built for stereoscopic content and consider little prior domain knowledge of the 3D human visual system (HVS) in network design. In this paper, to better simulate the binocular rivalry phenomenon, we propose a Predictive Auto-encoDing Network (PAD-Net) for blind/No-Reference stereoscopic image quality assessment (NR-SIQA). The proposed encoder-decoder architecture is inspired by the predictive coding theory that the cognition system tries to match bottom-up visual signal with top-down predictions. Besides, we design the Siamese framework to mimic the binocular rivalry in the 3D HVS based on the likelihood and prior maps generated from the predictive coding process. Extensive experimental results on three publicly available stereoscopic image quality databases demonstrate that the devised approach outperforms state-of-the-art algorithms for predicting the perceptual quality of both symmetrically and asymmetrically distorted stereoscopic images with various distortion types. Copyright © 2019, The Authors. All rights reserved.

关键词： Network coding

PUGAN: Physical Model-Guided Underwater Image Enhancement Using GAN with Dual-Discriminators

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Cong, Runmin Yang, Wenyu Zhang, Wei Li, Chongyi Guo, Chun-Le Huang, Qingming Kwong, Sam Institute of Information Science Beijing Jiaotong University Beijing100044 China School of Control Science and Engineering Shandong University Jinan250061 China Key Laboratory of Machine Intelligence and System Control Ministry of Education Jinan250061 China Beijing Key Laboratory of Advanced Information Science and Network Technology Beijing100044 China College of Computer Science Nankai University Tianjin300350 China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing101408 China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China Peng Cheng Laboratory Shenzhen518055 China Department of Computer Science City University of Hong Kong Hong Kong City University of Hong Kong Shenzhen Research Institute Shenzhen51800 China

Due to the light absorption and scattering induced by the water medium, underwater images usually suffer from some degradation problems, such as low contrast, color distortion, and blurring details, which aggravate the difficulty of downstream underwater understanding tasks. Therefore, how to obtain clear and visually pleasant images has become a common concern of people, and the task of underwater image enhancement (UIE) has also emerged as the times require. Among existing UIE methods, Generative Adversarial Networks (GANs) based methods perform well in visual aesthetics, while the physical model-based methods have better scene adaptability. Inheriting the advantages of the above two types of models, we propose a physical model-guided GAN model for UIE in this paper, referred to as PUGAN. The entire network is under the GAN architecture. On the one hand, we design a Parameters Estimation subnetwork (Par-subnet) to learn the parameters for physical model inversion, and use the generated color enhancement image as auxiliary information for the Two-Stream Interaction Enhancement subnetwork (TSIE-subnet). Meanwhile, we design a Degradation Quantization (DQ) module in TSIE-subnet to quantize scene degradation, thereby achieving reinforcing enhancement of key regions. On the other hand, we design the Dual-Discriminators for the style-content adversarial constraint, promoting the authenticity and visual aesthetics of the results. Extensive experiments on three benchmark datasets demonstrate that our PUGAN outperforms state-of-the-art methods in both qualitative and quantitative metrics. The code and results can be found from the link of https://***/proj_***. © 2023, CC BY-NC-SA.

关键词： Discriminators

Device-Edge Digital Semantic Communication with Trained Non-Linear Quantization

学校读者我要写书评

暂无评论

Device-Edge Digital Semantic Communication with Trained Non-...

IEEE Conference on Vehicular Technology (VTC)

作者： Lei Guo Wei Chen Yuxuan Sun Bo Ai State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University Beijing China Frontiers Science Center for Smart High-Speed Railway System School of Electronic and Information Engineering Beijing Jiaotong University Beijing China Key Laboratory of Railway Industry of Broadband Mobile Information Communications Beijing Engineering Research Center of High-Speed Railway Broadband Mobile Communications School of Information Engineering Zhengzhou University Zhengzhou China

Powered by deep learning, semantic communication is an intelligent communication paradigm, aiming to transmit useful information in the semantic domain. In most existing work, robust semantic features can be learned against wireless channel degradation, and directly transmitted in an analog fashion. However, analog semantic communication raises various challenges to the existing system from hardware/protocol to encryption issues. In this paper, we propose a novel non-linear quantization module to efficiently quantify semantic features. A sparse scaling vector is further incorporated to reduce the dimension of transmitted semantic features. Experimental results demonstrate that the proposed nonlinear quantization achieves better performance than the linear quantization method, and the performance of the digital system achieves better performance.

关键词：

On Approximate Opacity of Stochastic control systems

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Siyuan Yin, Xiang Dimarogonas, Dimos V. Zamani, Majid Division of Decision and Control Systems KTH Royal Institute of Technology Stockholm Sweden Department of Automation Shanghai Jiao Tong University Key Lab of System Control & Information Processing Ministry of Education Shanghai China Computer Science Department University of Colorado BoulderCO80309 United States Computer Science Department Ludwig Maximilian University of Munich Germany

This paper investigates an important class of information-flow security property called opacity for stochastic control systems. Opacity captures whether a system’s secret behavior (a subset of the system’s behavior that is considered to be critical) can be kept from outside observers. Existing works on opacity for control systems only provide a binary characterization of the system’s security level by determining whether the system is opaque or not. In this work, we introduce a quantifiable measure of opacity that considers the likelihood of satisfying opacity for stochastic control systems modeled as general Markov decision processes (gMDPs). We also propose verification methods tailored to the new notions of opacity for finite gMDPs by using value iteration techniques. Then, a new notion called approximate opacity-preserving stochastic simulation relation is proposed, which captures the distance between two systems’ behaviors in terms of preserving opacity. Based on this new system relation, we show that one can verify opacity for stochastic control systems using their abstractions (modeled as finite gMDPs). We also discuss how to construct such abstractions for a class of gMDPs under certain stability conditions. Copyright © 2024, The Authors. All rights reserved.

关键词： Stochastic control systems

LDGCN: An Edge-End Lightweight Dual GCN Based on Single-Channel EEG for Driver Drowsiness Monitoring

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Huang, Jingwei Wang, Chuansheng Huang, Jiayan Fan, Haoyi Grau, Antoni Zhang, Fuquan The College of Computer and Big Data Fuzhou University Fuzhou China The Department of Automatic Control Polytechnic University of Catalonia Barcelona Spain The New Engineering Industry College Putian University Putian China School of Computer and Artificial Intelligence Zhengzhou University China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Minjiang University Fuzhou China College of Computer and Control Engineering Minjiang University Fuzhou China Key Laboratory of Sichuan Province Sichuan Conservatory of Music Chengdu China Fuzhou Technology Innovation Center of Intelligent Manufacturing information System Minjiang University Fuzhou China Fujian Polytechnic Normal University Fujian Province University Fuzhou China

Driver drowsiness electroencephalography (EEG) signal monitoring can timely alert drivers of their drowsiness status, thereby reducing the probability of traffic accidents. Graph convolutional networks (GCNs) have shown significant advancements in processing the non-stationary, time-varying, and non-Euclidean nature of EEG signals. However, the existing single-channel EEG adjacency graph construction process lacks interpretability, which hinders the ability of GCNs to effectively extract adjacency graph features, thus affecting the performance of drowsiness monitoring. To address this issue, we propose an edge-end lightweight dual graph convolutional network (LDGCN). Specifically, we are the first to incorporate neurophysiological knowledge to design a Baseline Drowsiness Status Adjacency Graph (BDSAG), which characterizes driver drowsiness status. Additionally, to express more features within limited EEG data, we introduce the Augmented Graph-level Module (AGM). This module captures global and local information at the graph level, ensuring that BDSAG features remain intact while enhancing effective feature expression capability. Furthermore, to deploy our method on the fourth-generation Raspberry Pi, we utilize Adaptive Pruning Optimization (APO) on both channels and neurons, reducing inference latency by almost half. Experiments on benchmark datasets demonstrate that LDGCN offers the best trade-off between monitoring performance and hardware resource utilization compared to existing state-of-the-art algorithms. All our source code can be found at https://***/BryantDom/Driver-Drowsiness-Monitoring. Copyright © 2024, The Authors. All rights reserved.

关键词： Electroencephalography

Multi-Object Tracking Based on Anchor-Free Detection and Message Passing Association

学校读者我要写书评

暂无评论

Multi-Object Tracking Based on Anchor-Free Detection and Mes...

IEEE Conference on Industrial Electronics and Applications (ICIEA)

作者： Zhou Fang Zhi Zhang College of Computer Science and Technology Wuhan University of Science and Technology Wuhan China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System Wuhan China Big Data Science and Engineering Research Institute Wuhan University of Science and Technology Wuhan China

ISBN: (纸本)9781665409858

Due to the current multi-target tracking work, in the scene of dense and staggered targets, the tracked targets are prone to ID switch, which greatly affects the tracking effect. Aiming at this problem, this paper proposed a new multi-target tracking algorithm: using an improved anchor-free detection method and extracting features from multiple aspects, combined with the message passing graph neural network to participate in the data association process to further improve the exchange of feature information, so as to improve the algorithm's ability to associate target information data between frames and improve the tracking accuracy. After conducting quantitative analysis and comparison experiments on datasets such as MOT17 and MOT20, the experimental results show that the algorithm in this paper can effectively improve the problem of target ID switching and provide better tracking accuracy and performance.

关键词： Industrial electronics Target tracking Statistical analysis Message passing Switches Feature extraction Real-time systems