检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Liu, Hongda Wang, Yunlong Ren, Min Hu, Junxing Luo, Zhengquan Hou, Guangqi Sun, Zhenan The School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China The Center for Research on Intelligent Perception and Computing State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China The School of Artificial Intelligence Beijing Normal University Beijing100875 China The University of Science and Technology of China Hefei230027 China

Skeleton-based action recognition has recently made significant progress. However, data imbalance is still a great challenge in real-world scenarios. The performance of current action recognition algorithms declines sharply when training data suffers from heavy class imbalance. The imbalanced data actually degrades the representations learned by these methods and becomes the bottleneck for action recognition. How to learn unbiased representations from imbalanced action data is the key to long-tailed action recognition. In this paper, we propose a novel balanced representation learning method to address the long-tailed problem in action recognition. Firstly, a spatial-temporal action exploration strategy is presented to expand the sample space effectively, generating more valuable samples in a rebalanced manner. Secondly, we design a detached action-aware learning schedule to further mitigate the bias in the representation space. The schedule detaches the representation learning of tail classes from training and proposes an action-aware loss to impose more effective constraints. Additionally, a skip-modal representation is proposed to provide complementary structural information. The proposed method is validated on four skeleton datasets, NTU RGB+D 60, NTU RGB+D 120, NW-UCLA, and Kinetics. It not only achieves consistently large improvement compared to the state-of-the-art (SOTA) methods, but also demonstrates a superior generalization capacity through extensive experiments. Our code is available at https://***/firework8/BRL. Copyright © 2023, The Authors. All rights reserved.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

MACK: multimodal aligned conceptual knowledge for unpaired image-text matching 22

MACK: multimodal aligned conceptual knowledge for unpaired i...

引用

Proceedings of the 36th International Conference on Neural Information Processing systems

作者： Yan Huang Yuming Wang Yunan Zeng Liang Wang Center for Research on Intelligent Perception and Computing (CRIPAC) State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences (CASIA) School of Artificial Intelligence University of Chinese Academy of Sciences (UCAS) Chinese Academy of Sciences Artificial Intelligence Research (CAS-AIR)

ISBN: (纸本)9781713871088

Recently, the accuracy of image-text matching has been greatly improved by multimodal pretrained models, all of which are trained on millions or billions of paired images and texts. Different from them, this paper studies a new scenario as unpaired image-text matching, in which paired images and texts are assumed to be unavailable during model training. To deal with this, we propose a simple yet effective method namely Multimodal Aligned Conceptual Knowledge (MACK), which is inspired by the knowledge use in human brain. It can be directly used as general knowledge to correlate images and texts even without model training, or further fine-tuned based on unpaired images and texts to better generalize to certain datasets. In addition, we extend it as a re-ranking method, which can be easily combined with existing image-text matching models to substantially improve their performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A blockchain-based data transaction method with privacy protection and fairness

引用

Peer-to-Peer Networking and Applications 2025年第4期18卷 1-13页

作者： Yuan, Ye Yang, Mingxing Xiong, Ruoting Duan, Jia Liu, Xi Zhang, Lianchong Ren, Wei Zhang, Xin Hubei Land Resources Vocational College Wuhan China Gemological Institute China University of Geosciences Wuhan China School of Computing Science University of East Anglia Norwich UK The Third Surveying and Mapping Institute of Hunan Province Hunan China Hunan Engineering Research Center of Geographic Information Security and Application Hunan China Aerospace Information Research Institute Chinese Academy of Sciences Beijing China Provincial Key Laboratory of Multimodal Perceiving and Intelligent Systems Jiaxing University Jiaxing China School of Computer Science China University of Geosciences Wuhan China

In online trading, buyers may not be willing to reveal their identity, thus anonymous payment systems such as Central Bank Digital Currencies (CBDC) or blockchain systems may be preferred, instead of centralized payment systems. However, when relying on those payment systems, cheating or unfair behaviors from anonymous accounts should be defended against. In this paper, we propose a scheme that not only guarantees stronger customer privacy protection in terms of payment anonymity and transaction information anonymity (e.g., shopping user name, address), but also maintains the fairness in transactions. Our proposed system consists of three web systems: B2C web, M2B web, and delivery web, as well as an anonymous payment system (or blockchain, distributed ledger). The detailed models and protocols are proposed for trading, pre-payment, payment, arbitration between customers, merchants, manufacturers, third evaluators, and arbitrators. Experimental results show the low gas costs and execution time of each function on smart contract. The fairness in transactions is yet maintained because locked payment can be invoked by smart contract from oracle from the delivery web upon confirming the recipient of purchased merchandise by the customer.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Bootstrap Masked Visual Modeling via Hard Patches Mining

arXiv

引用

arXiv 2023年

作者： Wang, Haochen Fan, Junsong Wang, Yuxi Song, Kaiyou Wang, Tiancai Zhang, Xiangyu Zhang, Zhaoxiang The Center for Research on Intelligent Perception and Computing State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China The University of Chinese Academy of Sciences Beijing100190 China The Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science & Innovation Chinese Academy of Sciences 999077 Hong Kong Megvii Technology Beijing100096 China

Masked visual modeling has attracted much attention due to its promising potential in learning generalizable representations. Typical approaches urge models to predict specific contents of masked tokens, which can be intuitively considered as teaching a student (the model) to solve given problems (predicting masked contents). Under such settings, the performance is highly correlated with mask strategies (the difficulty of provided problems). We argue that it is equally important for the model to stand in the shoes of a teacher to produce challenging problems by itself. Intuitively, patches with high values of reconstruction loss can be regarded as hard samples, and masking those hard patches naturally becomes a demanding reconstruction task. To empower the model as a teacher, we propose Hard Patches Mining (HPM), predicting patch-wise losses and subsequently determining where to mask. Technically, we introduce an auxiliary loss predictor, which is trained with a relative objective to prevent overfitting to exact loss values. Also, to gradually guide the training procedure, we propose an easy-to-hard mask strategy. Empirically, HPM brings significant improvements under both image and video benchmarks. Interestingly, solely incorporating the extra loss prediction objective leads to better representations, verifying the efficacy of determining where is hard to reconstruct. The code is available at https://***/Haochen-Wang409/HPM. Copyright © 2023, The Authors. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Memory-Adaptive Vision-and-Language Navigation

SSRN

引用

SSRN 2023年

作者： He, Keji Jing, Ya Huang, Yan Lu, Zhihe An, Dong Wang, Liang Center for Research on Intelligent Perception and Computing State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China ByteDance AI Lab Beijing China National University of Singapore Singapore School of Future Technology University of Chinese Academy of Sciences Beijing China

Vision-and-Language Navigation (VLN) requests an agent to navigate in 3D environments following given instructions, where history is critical for decisionmaking in dynamic navigation process. Particularly, a memory bank storing histories is widely used in existing methods to incorporate with multimodel representations in current scenes for better decision-making. However, by weighting each history with a simple scalar, those methods cannot purely utilize the informative cues that co-exist with detrimental contents in each history, thereby inevitably introducing noises into decision-making. To that end, we propose a novel Memory-Adaptive Model (MAM) that can dynamically restrain the detrimental contents in histories for retaining contents that benefit navigation only. Specifically, two key modules, Visual and Textual Adaptive Modules, are designed to restrain history noises based on scene-related vision and text, respectively. A Reliability Estimator Module is further introduced to refine above adaptation operations. Our experiments on the widely used RxR and R2R datasets show that MAM outperforms its baseline method by 4.0% / 2.5% and 2% / 1% on the validation unseen/test split, respectively, wrt the SR metric. © 2023, The Authors. All rights reserved.

关键词： Navigation

来源：评论

学校读者我要写书评

暂无评论

Spin-Orbit Torque Driven Memristor in Bulk L10 FePt System for Neuromorphic computing

SSRN

引用

SSRN 2022年

作者： Tao, Ying Sun, Chao Li, Wendi Wang, Cen Jin, Fang Zhang, Yue Guo, Zhe Zheng, Yuhong Wang, Xiaoguang Dong, Kaifeng School of Automation China University of Geosciences Wuhan430074 China Hubei key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan430074 China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan430074 China Wuhan National Laboratory for Optoelectronics Huazhong University of Science & Technology Wuhan430074 China Changxin Memory Technologies Inc. Hefei230601 China

In this study, the memristor driven by spin-orbit torque (SOT) is realized in the bulk L1 0 FePt systems with high perpendicular magnetization anisotropy (PMA). Due to the domain nucleation and expansion driven by current pulses, multi-level Hall resistances can be continuously tuned by the current density, where the memristive states are retained by the domain wall pinning effects. The properties of multi-level resistances for samples with different structures are associated with the magnitude of field-like torque, and the larger efficiency of field-like torque enhances the multiple resistance characteristics. Furthermore, the memristive behaviors in the field-free condition are obtained by utilizing the interlayer exchange coupling. Finally, based on the memristive characteristics of our devices, the three-layer multilayer perceptron (MLP) neural network is built to perform the MNIST handwritten digit recognition task, and the accuracy of weight update can reach up to 88.55%. These results pave the way for the application of SOT-driven memristors in neuromorphic computing. © 2022, The Authors. All rights reserved.

关键词： Memristors

来源：评论

学校读者我要写书评

暂无评论

Identification of root cause alarms by detecting correlations and time delays between alarm variables

Identification of root cause alarms by detecting correlation...

引用

Chinese Automation Congress (CAC)

作者： Xin Cheng Wenkai Hu Weihua Cao Min Wu School of Automation China University of Geosciences Wuhan China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China

ISBN: (纸本)9781665426480

Under an abnormal situation in process operations, it is important to find out the root cause, so as to bring the system back to normal timely. In a typical modern industrial facility, there are configured a large number of alarms. Usually, the presence and propagation of a fault may lead to a series of alarms, which hold certain correlations. Therefore, analyzing alarm correlations can help to track the prorogation of faults among alarm variables and thus find out the root cause. This paper proposes a systematic alarm root cause analysis method based on the detection of correlations and time delays between alarm variables. First, two types of alarm correlations based on time stamps of alarm events are defined and the calculation methods are provided. Then, alarms are determined to be correlated or not by comparing with designed thresholds. Last, a causal matrix is formed to present the causal relations between alarms based on the directions of time delays. The effectiveness of the proposed method is demonstrated by both a numerical example and the benchmark Tennessee Eastman model.

关键词： Root cause analysis Correlation Systematics Automation Delay effects Benchmark testing Industrial facilities

来源：评论

学校读者我要写书评

暂无评论

Accelerometer optimization based on real-time measurement of Hole Deviation INS with While-Drilling

Accelerometer optimization based on real-time measurement of...

引用

Chinese Automation Congress (CAC)

作者： Yuan-Qiu Liu Fang Jin Jun-Lei Song Kai-Feng Dong Wen-Qin Mo Ya-Juan Hui School of Automation China University of Geosciences Wuhan China Hubei key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China

ISBN: (纸本)9781665426480

Well deviation Measurement of While-Drilling(MWD) refers to the real-time measurement of well deviation and azimuth While-Drilling is the key for navigation in automatic drilling, real-time detection of drilling direction and automatic drilling control. In this paper, we solve the problem that the accelerometer could not accurately measure gravity because of the high-speed rotation caused by the eccentric installation of INS (Inertial Navigation System) and the acceleration of vibration line in the real-time measurement of borehole deviation and azimuth angle while drilling. Firstly, the dynamic physical model of accelerometer and gyroscope under eccentric installation is built, the non-gravity error of accelerometer measurement under eccentric motion is analyzed, and the algorithm optimization is made. Secondly, a calibration scheme for the compensation coefficients of centrifugal and tangential forces is given, which enables the carrier to make eccentric pure rotational motion on the rotating platform to determine the compensation calibration coefficient matrix. Finally, the first-order complementary filter algorithm is used to fuse the gyroscope angular velocity with the corrected acceleration for attitude calculation, and the algorithm is experimented. In this study, MPU6050 is selected as the chip of INS, and the attitude algorithm is realized by STM32. The experimental results show that the high dynamic error of the accelerometer measurement caused by eccentric motion is effectively reduced by the algorithm compensation.

关键词： Accelerometers Drilling Semiconductor device measurement Heuristic algorithms Measurement uncertainty Vibration measurement Real-time systems

来源：评论

学校读者我要写书评

暂无评论

Embedding theory of reservoir computing and reducing reservoir network using time delays

引用

Physical Review Research 2023年第2期5卷 L022041-L022041页

作者： Xing-Yue Duan Xiong Ying Si-Yang Leng Jürgen Kurths Wei Lin Huan-Fei Ma School of Mathematical Sciences Soochow University Suzhou 215006 China School of Mathematical Sciences SCMS and SCAM Fudan University Shanghai 200433 China Research Institute of Intelligent Complex Systems and Centre for Computational Systems Biology Fudan University Shanghai 200433 China Academy for Engineering and Technology Fudan University Shanghai 200433 China State Key Laboratory of Medical Neurobiology LCNBI and MOE Frontiers Center for Brain Science Institutes of Brain Science Fudan University Shanghai 200032 China Potsdam Institute for Climate Impact Research (PIK) Potsdam 14473 Germany Shanghai Artificial Intelligence Laboratory Shanghai 200232 China

Reservoir computing (RC), a particular form of recurrent neural network, is under explosive development due to its exceptional efficacy and high performance in reconstruction and/or prediction of complex physical systems. However, the mechanism triggering such effective applications of RC is still unclear, awaiting deep and systematic exploration. Here, combining the delayed embedding theory with the generalized embedding theory, we rigorously prove that RC is essentially a high-dimensional embedding of the original input nonlinear dynamical system. Thus, using this embedding property, we unify into a universal framework the standard RC and the time-delayed RC where we introduce time delays only into the network's output layer, and we further find a trade-off relation between the time delays and the number of neurons in RC. Based on these findings, we significantly reduce the RC's network size and promote its memory capacity in completing systems reconstruction and prediction. More surprisingly, only using a single-neuron reservoir with time delays is sometimes sufficient for achieving reconstruction and prediction tasks, while the standard RC of any large size but without time delay cannot complete them yet.

关键词： Control & applications of chaos Mathematical physics Neuronal networks Patterns in complex systems

来源：评论

学校读者我要写书评

暂无评论

Speech Emotion Recognition Based on Deep Learning Network with Non-local Networks

Speech Emotion Recognition Based on Deep Learning Network wi...

引用

Chinese Automation Congress (CAC)

作者： M.T. Han Z.T. Liu School of Automation China University of Geosciences Wuhan China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China

ISBN: (纸本)9781665426480

Convolutional Neural Networks (CNN) based Speech Emotion Recognition (SER) has the problem of learning bias towards stronger emotions while giving less weight to relatively faint emotions. This lack of subtleness causes the misclassification of emotions that have similar arousal such as anger and joy. To solve this problem, a novel speaker-independent SER methods consist of CNN with non-local networks and attention-based Bi-directional Long Short-Term Memory (BiLSTM) is proposed. The non-local networks can capture the dependence between any two points globally to enhance the correlation of emotional features at the utterance level. The extracted trichannel Log-Mel spectrograms features include the raw of LogMel spectrums, the Log-Mel spectrums’ deltas and delta-deltas respectively. Finally long-term contextual dependencies, local information and non-local information which is obtained by nonlocal networks is contained in the extracted feature. It was tested on the Interactive Emotional Dyadic Motion Capture Database (IEMOCAP) dataset. The experimental results show that the WAR of SER is 66.85% and the UAR of SER is 61.01%. Compared with several newly proposed SER methods, the effectiveness of the proposed model is demonstrated.

关键词： Deep learning Emotion recognition Databases Speech recognition Feature extraction Data models Convolutional neural networks

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：