检索结果-内蒙古大学图书馆

TSPT: Two-Step Prompt Tuning for class-incremental novel class discovery

Knowledge-Based Systems 2025年 319卷

作者： An, Jiayu Du, Zhenbang Zhang, Herui Wu, Dongrui Key Laboratory of the Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen518063 China

In real-world applications, models often encounter a sequence of unlabeled new tasks, each containing unknown classes. This paper explores class-incremental novel class discovery (class-iNCD), which requires maintaining previously learned knowledge while discovering novel classes. We consider a more realistic and also more challenging scenario, which has a small number of initial known classes and a large number of unlabeled tasks, with the additional requirement of data privacy protection. A simple yet effective approach, Two-Step Prompt Tuning (TSPT), is proposed. TSPT tackles class-iNCD through prompt tuning, which is rehearsal-free and plug-and-play, protecting data privacy and significantly reducing the number of trainable parameters. TSPT consists of two main steps: (1) novel class discovery, which initializes the classifier using uniform clustering, and uses intra- and inter-sample consistency learning to discover novel classes;and, (2) knowledge fusion, where the prompt learned in the previous step is adapted as task-specific prompt, and additional optimal prompts are selected from a prompt pool to integrate knowledge from both old and new classes. Experiments on three datasets demonstrated the effectiveness of TSPT. © 2025 Elsevier B.V.

关键词： Sensitive data

来源：评论

学校读者我要写书评

暂无评论

Surrogate-Assisted Multiobjective Gene Selection for Cell Classification from Large-Scale Single-Cell RNA Sequencing Data

引用

IEEE Transactions on Evolutionary Computation 2025年第3期29卷 601-615页

作者： Lin, Jianqing He, Cheng Jiang, Hanjing Huang, Yabing Jin, Yaochu Huazhong University of Science and Technology Key Laboratory of Image Information Processing and Intelligent Control of Education Ministry of China School of Artificial Intelligence and Automation Wuhan430074 China Huazhong University of Science and Technology State Key Laboratory of Advanced Electromagnetic Technology School of Electrical and Electronic Engineering Wuhan430074 China Renmin Hospital Department of Pathology Wuhan University Wuhan430060 China Westlake University Trustworth and General Artificial Intelligence Laboratory School of Engineering Hangzhou310030 China

Accurate cell classification is crucial but expensive for large-scale single-cell RNA sequencing (scRNA-seq) analysis. Gene selection (GS) emerges as a pivotal technique in identifying gene subsets of scRNA-seq for classification accuracy improvement and gene scale reduction. Nevertheless, the rising scale of scRNA-seq data presents challenges to existing GS methods regarding performance and computational time. Thus, we propose a surrogate-assisted evolutionary algorithm for multiobjective GS to address these deficiencies. An innovative two-phase initialization method is proposed to select sparse solutions to provide preliminary insights into gene contributions. Then, a binary competitive swarm optimizer is proposed for effective global search, where a local search method is embedded to eliminate irrelevant genes for efficiency consideration. Additionally, a surrogate model is adopted to forecast classification accuracy efficiently and substitutes part of the computationally expensive classification process. Experiments are conducted on eight large-scale scRNA-seq datasets with more than 20,000 genes. The effectiveness of the proposed GS method for scRNA-seq cell classification compared with eight state-of-the-art methods is validated. Gene expression analysis results of selected genes further validated the significance of the genes selected by the proposed method in the classification of scRNA-seq data. © 1997-2012 IEEE.

关键词： Gene expression

来源：评论

学校读者我要写书评

暂无评论

Prescribed Performance Control of Two-Time-Scale Systems and Its Application

引用

International Journal of Robust and Nonlinear Control 2025年

作者： Zeng, Ze-Hong Wang, Yan-Wu Liu, Xiao-Kang Yang, Wu School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China Key Laboratory of Image Processing and Intelligent Control Huazhong University of Science and Technology Ministry of Education Wuhan China School of Electrical Engineering Guangxi University Nanning China

This article explores the prescribed performance control design for linear two-time-scale systems (TTSSs). Due to ill-conditioning and high dimensionality, existing prescribed performance control methods for single-time-scale systems are unsuitable for TTSSs. Moreover, current TTSS methodologies focus on steady-state performance, often neglecting transient dynamics. To address these challenges, we first apply the Chang transformation to decouple the fast and slow states. Next, we use a state transformation to convert the state equation into block form to eliminate the requirement of a full row-rank input matrix. Finally, the backstepping method is utilized to design the prescribed performance control. The effectiveness and advantages of the proposed control strategy are demonstrated through two examples: a numerical simulation and a hardware-in-the-loop experiment involving an electronic circuit system. © 2025 John Wiley & Sons Ltd.

关键词： backstepping control prescribed performance control two-time-scale system underactuated system

来源：评论

学校读者我要写书评

暂无评论

A Parallel Read-Write Circuit With Fast Amplitude-Adaptive Matching Scheme to Memristor Crossbar Array

引用

IEEE Transactions on Circuits and Systems I: Regular Papers 2025年

作者： Jiang, Ningye Jiang, Mingxuan Huang, Haoen Yang, Le Huang, Yi Zeng, Zhigang Huazhong University of Science and Technology School of Artificial Intelligence and Automation Key Laboratory of Image Information Processing and Intelligent Control Ministry of Education of China Wuhan430074 China Wuhan Institute of Technology School of Electrical and Information Engineering Wuhan430205 China University of Massachusetts Amherst Department of Electrical and Computer Engineering AmherstMA01003 United States

Memristor crossbar array (MCA) is a computing-in-memory (CIM) module for computational acceleration. However, conventional read-write (R-W) circuits for MCA rely heavily on external components and have shortcoming in long writing times. To address these issues, we propose a parallel R-W circuit with a fast amplitude-adaptive matching (AAM) scheme. The fast AAM scheme is designed to accelerate time of writing memristors in MCA by adaptively adjusting the writing voltage to match an optimal amplitude. The architecture of parallel R-W circuit are outlined with the implementation of R-W processes. The design principles of parallel R-W unit, fast AAM unit, and associated control logic are described for analyzing their functions within the circuit. By integrating the parallel R-W circuit with MCA, a computing block is leveraged in a neural network for classification tasks. The experimental results demonstrate that the neural network with multiple computing blocks maintains high accuracy across various datasets (over 80%). Furthermore, the proposed AAM scheme achieves an average 54% reduction in writing time compared to that one without AAM scheme. Compared to the pulse-width scheme, the parallel R-W circuit exhibits an average 3.05× speedup in adjusting a 5×5 MCA, and the speedup efficiency is higher as the increases of MCA size. © 2025 IEEE.

关键词： Memristors

来源：评论

学校读者我要写书评

暂无评论

Dynamics of Masked image Modeling in Hyperspectral image Classification

引用

IEEE Transactions on Geoscience and Remote Sensing 2025年 63卷

作者： Ma, Chen Li, Huayi Jiang, Junjun Aybar, Cesar Yao, Jiaqi Camps-Valls, Gustau Harbin Institute of Technology Research Center of Satellite Technology State Key Laboratory of Micro-Spacecraft Rapid Design and Intelligent Cluster School of Astronautics Harbin150001 China Harbin Institute of Technology School of Computer Science and Technology Harbin150001 China University of València Image Signal Processing Laboratory València46980 Spain Tianjin Normal University Tianjin300387 China

Masked image modeling (MIM), a common self-supervised learning (SSL) technique, has been extensively studied for remote sensing (RS) image processing. Nevertheless, its effectiveness for hyperspectral imagery (HSI) remains underexplored due to the distinct data structures and high dimensionality. This article aims to provide a detailed understanding of MIM from different perspectives of representation learning and statistical analysis for HSI classification tasks. Our study reveals that the MIM paradigm injects inductive bias in the attention mechanism of the transformer model, which is advantageous for capturing the local discrepancies between the spectra. We also show that MIM can increase the diversity of the attention heads in every layer, which is beneficial for the model in extracting more discriminative features from different spectral bands. The similarity of representations from various layers further proves this. Furthermore, our investigation highlights how MIM introduces a dynamic perspective to spectral representations, enabling the model to learn more robust and discriminative features. The final numerical experiments indicate that a moderate mask ratio can enhance the performance of downstream tasks. This suggests that designing a more targeted masking strategy might be necessary to achieve higher and more stable gains in downstream task performance. Without bells and whistles, the vanilla MIM improves the overall classification accuracy by an average of 2.69% over its supervised learning (SL) counterpart. We hope that our findings can advance the understanding of MIM in HSI and inspire the design of a more stable SSL paradigm for HSI processing. © 1980-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Deep transfer learning-based decoder calibration for intracortical brain-machine interfaces

引用

Computers in Biology and Medicine 2025年第Pt A期192卷 110231-110231页

作者： Li, Xiao Dong, Xianxin Wang, Jun Mao, Haodong Tu, Xikai Li, Wei He, Jiping Li, Qiang Zhang, Peng Hubei Key Laboratory of Modern Manufacturing Quantity Engineering School of Mechanical Engineering Hubei University of Technology Wuhan430068 China Britton Chance Center for Biomedical Photonics Wuhan National Laboratory for Optoelectronics Huazhong University of Science and Technology Hubei Wuhan China MoE Key Laboratory for Biomedical Photonics Collaborative Innovation Center for Biomedical Engineering School of Engineering Sciences Huazhong University of Science and Technology Hubei Wuhan China The Key Laboratory of Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Hubei Wuhan China Advanced Innovation Center for Intelligent Robots and Systems Beijing Institute of Technology Beijing China

Intracortical brain-machine interfaces (iBMIs) aim to establish a communication path between the brain and external devices. However, in the daily use of iBMIs, the non-stationarity of recorded neural signals necessitates frequent recalibration of the iBMI decoder to maintain decoding performance, which requires collecting and labeling a large amount of new data. To address this challenge and minimize the time needed for decoder recalibration, we proposed an active learning domain adversarial neural network (AL-DANN). This model leveraged a substantial volume of historical data alongside a small amount of current data (four samples per category) to calibrate the decoder. By incorporating domain adversarial and active learning strategies, the model effectively transferred knowledge from historical data to new data, reducing the demand for new samples. We validated the proposed method using neural signals recorded from three monkeys performing different movements in a classification task or a regression task. The results showed that the AL-DANN outperformed existing state-of-the-art methods. Impressively, it required only four new samples per category for decoder recalibration, leading to a recalibration time reduction of over 80 %. To our knowledge, this is the first study to incorporate deep transfer learning into iBMI decoder calibration, highlighting the significant potential of applying deep learning technologies in iBMIs. © 2025

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Adaptively wavelet-based image denoising algorithm with edge preserving

引用

Chinese Optics Letters 2006年第2期4卷 80-83页

作者：谭毅华田金文柳健 State Key Laboratory of Education Ministry for Image Processing and Intelligent Control

A new wavelet-based image denoising algorithm, which exploits the edge information hidden in the corrupted image, is presented. Firstly, a canny-like edge detector identifies the edges in each subband. Secondly, multiplying the wavelet coefficients in neighboring scales is implemented to suppress the noise while magnifying the edge information, and the result is utilized to exclude the fake edges. The isolated edge pixel is also identified as noise. Unlike the thresholding method, after that we use local window filter in the wavelet domain to remove noise in which the variance estimation is elaborated to utilize the edge intbrmation. This method is adaptive to local image details, and can achieve bet, ter performance than the methods of state of the art.

关键词： Adaptive algorithms Detectors Edge detection Information technology Noise abatement Pattern recognition Wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

Air Pollution Monitoring By Integrating Local and Global Information in Self-Adaptive Multiscale Transform Domain

引用

IEEE Transactions on Multimedia 2025年

作者： Gu, Ke Liu, Yuchen Liu, Hongyan Liu, Bo Qiao, Junfei Lin, Weisi Zhang, Wenjun Beijing University of Technology School of Information Science and Technology Engineering Research Center of Intelligent Perception and Autonomous Control of Ministry of Education Beijing Laboratory of Smart Environmental Protection Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing Artificial Intelligence Institute Beijing100124 China Massey University School of Mathematical and Computational Sciences New Zealand Nanyang Technological University College of Computing and Data Science Singapore Shanghai Jiao Tong University Institute of Image Communication and Information Processing Shanghai China

As the deadliest form of pollution, air pollution had a prolonged severe damage to the human health and life safety of nearly 99% of the world's population. Facing to the problem that billions of tons of pollutants discharged into the air per year by industrial activities, this paper proposed a novel image-based air pollution monitor (IAPM) by incorporating local and global information in the self-adaptive multiscale transform domain, so as to achieve the timely and effective leakage detection of typical air pollutants from a single image. To be specific, this paper first developed a screen-shaped module according to two significant findings in visual neuroscience, which include the high sensitivity of human eyes to horizontal and vertical stimuli and the center-surround inhibition, by designing and fusing the square module, horizontal strip module and vertical strip module parallelly for simulating the behaviour of human eyes to extract local features. Second, the learnable weights and proportional mapping were applied to incorporate the screen-shaped module and lightweight vision transformer as backbone, towards more richly exploiting and fusing local and global information just as the way a brain perceives external stimuli. Third, a new self-adaptive multiscale transform domain method was devised based on two motivations from the visual characteristics of multiscale perception and the brain characteristics of self-adaptive domain transform to modify the backbone by using the operations of pooling and pointwise convolution. Extensive experiments implemented on the datasets of carbon particulate matters and ethylene leakage confirmed the superior monitoring performance of the proposed IAPM model beyond the state-of-the-art (SOTA) peers by an accuracy gain of about 4%. Furthermore, the proposed IAPM model only required 0.089 GFLOPs and 0.15 million model parameters, remarkably outperforming SOTA competitors in computational efficiency and storage resources. © 1999-2

关键词： Information leakage

来源：评论

学校读者我要写书评

暂无评论

Data alignment based adversarial defense benchmark for EEG-based BCIs

引用

Neural Networks 2025年 188卷 107516页

作者： Chen, Xiaoqing Jia, Tianwang Wu, Dongrui Key Laboratory of the Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen518063 China Zhongguancun Academy Beijing100080 China

Machine learning has been extensively applied to signal decoding in electroencephalogram (EEG)-based brain–computer interfaces (BCIs). While most studies have focused on enhancing the accuracy of EEG-based BCIs, more attention should be given to their security. Recent findings reveal that EEG-based BCIs are vulnerable to adversarial attacks. To address this, we present the first adversarial defense benchmark based on data alignment, aiming to enhance both the accuracy and robustness of EEG-based BCIs. This study evaluates nine adversarial defense approaches (including five defense strategies) across five EEG datasets (covering three paradigms), three neural networks, and four experimental scenarios. Our results show for the first time that integrating data augmentation, data alignment, and robust training can further improve both the accuracy and robustness of BCIs compared to using only one or two of them. Furthermore, we provide insights into the characteristics of various adversarial defense approaches based on EEG data alignment, offering valuable guidance for developing more accurate and secure EEG-based BCIs. © 2025 Elsevier Ltd

关键词： Adversarial attack Adversarial defense Brain–computer interface Data alignment Electroencephalogram Security

来源：评论

学校读者我要写书评

暂无评论

Spiking neural P systems with anti-spikes and without annihilating priority as number acceptors

引用

Journal of Systems Engineering and Electronics 2014年第3期25卷 464-469页

作者： Gangjun Tan Tao Song Zhihua Chen Key Laboratory of Image Information Processing and Intelligent Control School of AutomationHuazhong University of Science and Technology

Spiking neural P systems with anti-spikes （ASN P systems） are variant forms of spiking neural P systems, which are inspired by inhibitory impulses/spikes or inhibitory synapses. The typical feature of ASN P systems is when a neuron contains both spikes and anti-spikes, spikes and anti-spikes wil immediately annihilate each other in a maximal way. In this paper, a restricted variant of ASN P systems, cal ed ASN P systems without anni-hilating priority, is considered, where the annihilating rule is used as the standard rule, i.e., it is not obligatory to use in the neuron associated with both spikes and anti-spikes. If the annihilating rule is used in a neuron, the annihilation wil consume one time unit. As a result, such systems using two categories of spiking rules （identified by （a, a） and （a,a^-）） can achieve Turing completeness as number accepting devices.

关键词： membrane computing spiking neural P system Turing completeness register machine anti-spike.

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

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

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：