检索结果-内蒙古大学图书馆

Forget less,count better:a domain-incremental self-distillation learning benchmark for lifelong crowd counting

Frontiers of Information Technology & Electronic Engineering 2023年第2期24卷 187-202页

作者： Jiaqi GAO Jingqi LI Hongming SHAN Yanyun QU James ZWANG Fei-Yue WANG Junping ZHANG Shanghai Key Laboratory of Intelligent Information Processing School of Computer ScienceFudan UniversityShanghai 200433China Institute of Science and Technology for Brain-inspired Intelligence Fudan UniversityShanghai 200433China Shanghai Center for Brain Science and Brain-inspired Technology Shanghai 201210China School of Information Science and Technology Xiamen UniversityXiamen 361005China College of Information Sciences and Technology the Pennsylvania State UniversityUniversity ParkPA 16802USA State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190China

Crowd counting has important applications in public safety and pandemic control.A robust and practical crowd counting system has to be capable of continuously learning with the newly incoming domain data in real-world scenarios instead of fitting one domain ***-the-shelf methods have some drawbacks when handling multiple domains:(1)the models will achieve limited performance(even drop dramatically)among old domains after training images from new domains due to the discrepancies in intrinsic data distributions from various domains,which is called catastrophic forgetting;(2)the well-trained model in a specific domain achieves imperfect performance among other unseen domains because of domain shift;(3)it leads to linearly increasing storage overhead,either mixing all the data for training or simply training dozens of separate models for different domains when new ones are *** overcome these issues,we investigate a new crowd counting task in incremental domain training setting called lifelong crowd *** goal is to alleviate catastrophic forgetting and improve the generalization ability using a single model updated by the incremental ***,we propose a self-distillation learning framework as a benchmark(forget less,count better,or FLCB)for lifelong crowd counting,which helps the model leverage previous meaningful knowledge in a sustainable manner for better crowd counting to mitigate the forgetting when new data arrive.A new quantitative metric,normalized Backward Transfer(nBwT),is developed to evaluate the forgetting degree of the model in the lifelong learning *** experimental results demonstrate the superiority of our proposed benchmark in achieving a low catastrophic forgetting degree and strong generalization ability.

关键词： Crowd counting Knowledge distillation Lifelong learning

来源：评论

学校读者我要写书评

暂无评论

Fluorescence image-guided tumour surgery

NATURE REVIEWS BIOENGINEERING

引用

NATURE REVIEWS BIOENGINEERING 2023年第3期1卷 161-179页

作者： Wang, Kun Du, Yang Zhang, Zeyu He, Kunshan Cheng, Zhongquan Yin, Lin Dong, Di Li, Changjian Li, Wei Hu, Zhenhua Zhang, Chong Hui, Hui Chi, Chongwei Tian, Jie Chinese Acad Sci Inst Automat CAS Key Lab Mol Imaging Beijing Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing Peoples R China Chinese Acad Sci Inst Automat Beijing Key Lab Mol Imaging Beijing Peoples R China Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing Peoples R China Beihang Univ Beijing Adv Innovat Ctr Big Data Based Precis Med Sch Engn Med Beijing Peoples R China Chinese Acad Sci Inst Software State Key Lab Comp Sci Beijing Peoples R China Chinese Acad Sci Inst Software Beijing Key Lab Human Comp Interact Beijing Peoples R China Capital Med Univ Beijing Friendship Hosp Dept Gen Surg Beijing Peoples R China Jinan Univ Affiliated Hosp 1 Med Imaging Ctr Guangzhou Peoples R China Beijing Technol & Business Univ Sch Int Econ & Management Dept Big Data Management & Applicat Beijing Peoples R China

Intraoperative fluorescent imaging can provide real-time identification of tumours, lymph nodes, nerves and other healthy and malignant tissues during oncological surgery, contributing to better surgical outcomes. Various fluorescent probes have been clinically approved for surgical applications, improving tumour resection precision and preventing iatrogenic injury. In this Review, we discuss the development and application of fluorescent probes for image-guided surgery, including systemically and locally applied probes that are either non-targeted or targeted to specific tumours and tissues. We discuss the optimization and clinical potential of these probes, and highlight their current and future applications in oncological surgery. In addition, we examine the hardware of fluorescence imaging equipment, and discuss how artificial intelligence can enable real-time quantification to guide surgical decision-making. Finally, we highlight the remaining challenges in the field of image-guided surgery, including the need for standardization to achieve regulatory approval of new probes, and the required team effort for new probe development.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Application of machine learning method in optical molecular imaging: a review

引用

Science China(Information Sciences) 2020年第1期63卷 5-20页

作者： Yu AN Hui MENG Yuan GAO Tong TONG Chong ZHANG Kun WANG Jie TIAN CAS Key Laboratory of Molecular Imaging Beijing Key Laboratory of Molecular ImagingThe State Key Laboratory of Management and Control for Complex SystemsInstitute of Automation Chinese Academy of Sciences Beijing Advanced Innovation Center for Big Data-Based Precision Medicine School of MedicineBeihang University University of Chinese Academy of Sciences

Optical molecular imaging(OMI) is an imaging technology that uses an optical signal, such as near-infrared light, to detect biological tissue in organisms. Because of its specific and sensitive imaging performance, it is applied in both preclinical research and clinical surgery. However, it requires heavy data analysis and a complex mathematical model of tomographic imaging. In recent years, machine learning(ML)-based artificial intelligence has been used in different fields because of its ability to perform powerful data processing. Its analytical capability for processing complex and large data provides a feasible scheme for the requirement of OMI. In this paper, we review ML-based methods applied in different OMI modalities.

关键词： optical molecular imaging machine learning artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Generalized Zero-Shot Learning via Implicit Attribute Composition

Generalized Zero-Shot Learning via Implicit Attribute Compos...

引用

IEEE International Conference on systems, Man and Cybernetics

作者： Lei Zhou Yang Liu Qiang Li School of Computer Science and Artificial Intelligence Wuhan University of Technology Wuhan China College of Computer Science and Technology Zhejiang University Hangzhou China State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

Zero-shot learning (ZSL) is an important but challenging task in computer vision that aims to identify unseen classes without matching training samples. Current cutting-edge ZSL methods based on locality focus on acquiring the explicit locality of distinguishing characteristics, which could face a lack of adequate supervision at the class attribute level. This paper introduces a novel approach called IAC, which aims to learn Implicit Attribute Composition for ZSL. This method is more comprehensive compared to attribute localization that solely focuses on class-level attribute supervision. IAC utilizes subspace representations that efficiently capture the inherent structure of high-dimensional image features. Then, we learn implicit attribute composition through subspace representation learning. The superiority of the proposed IAC compared to the state-of-the-art is demonstrated through sufficient experiments conducted on three commonly used ZSL datasets, CUB, SUN, and AwA2.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Spatial-Temporal Attention Model for Human Trajectory Prediction

引用

IEEE/CAA Journal of Automatica Sinica 2020年第4期7卷 965-974页

作者： Xiaodong Zhao Yaran Chen Jin Guo Dongbin Zhao the School of Automation and Electrical Engineering University of Science and Technology BeijingBeijing 100083 the State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190China the Key Laboratory of Knowledge Automation for Industrial Processes Ministry of EducationBeijing 100083China

Human trajectory prediction is essential and promising in many related applications. This is challenging due to the uncertainty of human behaviors, which can be influenced not only by himself, but also by the surrounding environment. Recent works based on long-short term memory(LSTM) models have brought tremendous improvements on the task of trajectory prediction. However, most of them focus on the spatial influence of humans but ignore the temporal influence. In this paper, we propose a novel spatial-temporal attention(ST-Attention) model,which studies spatial and temporal affinities jointly. Specifically,we introduce an attention mechanism to extract temporal affinity,learning the importance for historical trajectory information at different time instants. To explore spatial affinity, a deep neural network is employed to measure different importance of the neighbors. Experimental results show that our method achieves competitive performance compared with state-of-the-art methods on publicly available datasets.

关键词： Attention mechanism long-short term memory(LSTM) spatial-temporal model trajectory prediction

来源：评论

学校读者我要写书评

暂无评论

Neurons Perception Dataset for RoboMaster AI Challenge

Neurons Perception Dataset for RoboMaster AI Challenge

引用

IEEE International Conference on Fuzzy systems (FUZZ-IEEE) / IEEE World Congress on Computational Intelligence (IEEE WCCI) / International Joint Conference on Neural Networks (IJCNN) / IEEE Congress on Evolutionary Computation (IEEE CEC)

作者： Li, Haoran Duan, Zicheng Li, Jiaqi Ma, Mingjun Chen, Yaran Zhao, Dongbin Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing 100190 Peoples R China Univ Chinese Acad Sci Beijing Peoples R China Australian Natl Univ Coll Engn & Comp Sci Canberra ACT 2601 Australia Beijing Inst Technol Sch Mech Engn Beijing 100081 Peoples R China

ISBN: (数字)9781728186719

ISBN: (纸本)9781728186719

From virtual game to physical robot, games have witnessed the development of artificial intelligence (AI) technology, especially the data-driven technology represented by deep learning. Compared with virtual games, a physical robot game such as RoboMaster AI challenge needs to build a complete closed-loop architecture composed of perception, planning, control, and decision-making to support autonomous confrontation. Perception, as the eye of the robot, its performance in the complex environment depends on a massive dataset. Although there are many open perception datasets, these datasets are difficult to meet the needs of RoboMaster AI challenge due to the high dynamics of the task, the distinctiveness of the objects, and limited computing resources. In this paper, we release a dataset named Neurons 1 perception dataset for RoboMaster AI challenge, which covers 3 tasks including monocular depth estimation, lightweight object detection, and multi-view 3D object detection, and makes up the data blank in this field. In addition, we also evaluate state-Of-The-Art (SOTA) methods on each task, hoping to provide an impartial benchmark for the development of perception algorithm.

关键词： Deep learning Three-dimensional displays Estimation Object detection Games Benchmark testing Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

AGI in Metaverse for Smart Cities and Societies: A Cyber Physical Social Approach

AGI in Metaverse for Smart Cities and Societies: A Cyber Phy...

引用

control Conference (ANZCC), Australian and New Zealand

作者： Xiao Wang Jun Huang Yong-Lin Tian Fei-Yue Wang Engineering Research Center of Autonomous Unmanned System Technology Ministry of Education Anhui University Hefei China Faculty of Innovation Engineering Macau University of Science and Technology Macao China State Key Laboratory for Management and Control of Complex Systems Chinese Academy of Sciences Beijing China Macao Institute of Systems Engineering Macau University of Science and Technology Macao China

The development of artificial general intelligence (AGI) and metaverse technologies hold great potential to enhance the effectiveness of smart cities and societies. we present a cyberphysical-social approach to integrating AGI into metaverse-based smart cities and societies, with a specific focus on parallel intelligence. The proposed approach considers the interplay between the physical environment, digital systems, social dynamics, and emerging forms of parallel intelligence, as well as swarm intelligence and collective intelligence. By leveraging these forms of parallel intelligence, the integration of AGI can create more responsive and proactive infrastructure, personalized and efficient services, and foster greater community engagement. The metaverse acts as a platform for simulating real-world scenarios and testing and refining AGI models before deployment in the physical world, while also facilitating the development and coordination of parallel intelligence. However, ethical considerations must be taken into account to ensure responsible deployment and avoid exacerbating existing societal inequalities or infringing upon individual privacy rights. This report provides a comprehensive framework for the integration of AGI and parallel intelligence into metaverse-based digital and smart organizations, while emphasizing the importance of human ethics, social responsibility, and ecological sustainability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Traffic-Condition-Awareness Ensemble Learning for Traffic Flow Prediction 3rd

Traffic-Condition-Awareness Ensemble Learning for Traffic Fl...

引用

3rd IFAC Workshop pn Cyber-Physical and Human systems (CPHS)

作者： Chen, Yuanyuan Lv, Yisheng Ye, Peijun Zhu, Fenghua Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing 100190 Peoples R China

Traffic prediction is an elemental function of Intelligent Transportation systems (ITS), and accurate and timely prediction is significant for both proactive traffic control and providing traveler information. In this paper, we focus on investigating ensemble leaning that benefits from different base models, and propose a traffic condition-awareness ensemble approach. We apply graph convolution on the network of traffic detectors to capture the spatial patterns embedded in traffic flow. Then, the extracted features are used to formulate a weight matrix to ensemble the predictions of base models according to their performances under a certain condition. We performed a series of experiments on a real dataset to compare the proposed methods with several competitive models, including ensemble methods: Weight Regression model and Gradient Boosting Regression Tree model, and single model approach: Support Vector Regression (SVR), Long Short-term Memory (LSTM) model and Historical Average Model model. Experimental results demonstrate that our method can significantly improve the performances of traffic flow prediction. Copyright (C) 2020 The Authors.

关键词： Intelligent road transportation Urban mobility

来源：评论

学校读者我要写书评

暂无评论

POLSAR IMAGE FEATURE EXTRACTION BASED ON CO-REGULARIZATION

POLSAR IMAGE FEATURE EXTRACTION BASED ON CO-REGULARIZATION

引用

IEEE International Geoscience and Remote Sensing Symposium (IGARSS)

作者： Huang, Xiayuan Nie, Xiangli Qiao, Hong Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing 100190 Peoples R China

ISBN: (纸本)9781728163741

Wishart distance of covariance matrices and Euclidean distance of polarimetric features are two important similarity measurements for polarimetric synthetic aperture radar (PolSAR) image classification. This paper proposes a feature extraction method by combing the two distances for PolSAR image classification. Firstly, two weight graphs are constructed based on the two distances to represent the local information of the data. Specifically, the neighbouring samples are sought in a local region to reduce the computation burden and utilize the spatial information. Then the dimensionality reduction model is constructed based on the two weight graphs and co-regularization. The co-regularization aims to minimize the dissimilarity of low-dimensional features corresponding to two graphs. Finally, the obtained low-dimensional features are used for Po1SAR image classification. Experimental results on real PolSAR datasets demonstrate the effectiveness and superiority of the proposed method.

关键词： Wishart distance Euclidean distance of polarimetric features dimensionality reduction PolSAR image classification feature extraction

来源：评论

学校读者我要写书评

暂无评论

Mixed Motivation Driven Social Multi-Agent Reinforcement Learning for Autonomous Driving

引用

IEEE/CAA Journal of Automatica Sinica 2025年第6期12卷 1272-1282页

作者： Long Chen Peng Deng Lingxi Li Xuemin Hu State Key Laboratory of Multimodal Artificial Intelligence Systems and the State Key Laboratory of Management and Control for Complex Systems Chinese Academy of Sciences Beijing WAYTOUS Inc. Beijing Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ) Shenzhen Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University Xi'an China School of Computer Science and Information Engineering Hubei University Wuhan China Purdue School of Engineering and Technology Indiana University-Purdue University Indianapolis Indianapolis IN USA School of Artificial Intelligence Hubei University Wuhan Key Laboratory of Intelligent Sensing System and Security (Hubei University) Ministry of Education Wuhan China

Despite great achievement has been made in autonomous driving technologies, autonomous vehicles (AVs) still exhibit limitations in intelligence and lack social coordination, which is primarily attributed to their reliance on single-agent technologies, neglecting inter-AV interactions. Current research on multi-agent autonomous driving (MAAD) predominantly focuses on either distributed individual learning or centralized cooperative learning, ignoring the mixed-motive nature of MAAD systems, where each agent is not only self-interested in reaching its own destination but also needs to coordinate with other traffic participants to enhance efficiency and safety. Inspired by the mixed motivation of human driving behavior and their learning process, we propose a novel mixed motivation driven social multi-agent reinforcement learning method for autonomous driving. In our method, a multi-agent reinforcement learning (MARL) algorithm, called Social Learning Policy Optimization (SoLPO), which takes advantage of both the individual and social learning paradigms, is proposed to empower agents to rapidly acquire self-interested policies and effectively learn socially coordinated behavior. Based on the proposed SoLPO, we further develop a mixed-motive MARL method for autonomous driving combined with a social reward integration module that can model the mixed-motive nature of MAAD systems by integrating individual and neighbor rewards into a social learning objective for improved learning speed and effectiveness. Experiments conducted on the MetaDrive simulator show that our proposed method outperforms existing state-of-the-art MARL approaches in metrics including the success rate, safety, and efficiency. More-over, the AVs trained by our method form coordinated social norms and exhibit human-like driving behavior, demonstrating a high degree of social coordination.

关键词： Measurement Reinforcement learning Safety Driver behavior Autonomous vehicles Optimization

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：