检索结果-内蒙古大学图书馆

HSI-MFormer: Integrating Mamba and Transformer Experts for Hyperspectral Image Classification

IEEE Transactions on Geoscience and Remote Sensing 2025年 63卷

作者： He, Yan Tu, Bing Liu, Bo Li, Jun Plaza, Antonio Institute of Optics and Electronics State Key Laboratory Cultivation Base of Atmospheric Optoelectronic Detection and Information Fusion Jiangsu International Joint Laboratory on Meteorological Photonics and Optoelectronic Detection Jiangsu Engineering Research Center for Intelligent Optoelectronic Sensing Technology of Atmosphere Nanjing University of Information Science and Technology Nanjing210044 China China University of Geosciences Faculty of Computer Science Wuhan430074 China University of Extremadura Hyperspectral Computing Laboratory Department of Technology of Computers and Communications Escuela Politecnica 10003 Spain

Hyperspectral image (HSI) classification is fundamental to numerous remote sensing applications, enabling detailed analysis of material properties and environmental conditions. Recent Mamba built upon selective state space models (SSMs) (S6) have demonstrated exceptional advantages in long-range sequence modeling with linear computational efficiency, while Transformer based on self-attention mechanisms is particularly adept at capturing short-range dependencies. To leverage the complementary strengths of these models, this article introduces a novel hybrid Mamba-Transformer framework (HSI-MFormer), effectively exploring the multiscale properties of hyperspectral data for HSI classification. Initially, a multiscale token generation (MTG) module is developed, which converts the HSI cube into multiple spatial-spectral token groups across different scales. To adequately capture fine-grained multiscale spatial-spectral patterns, an inner-scale transformer expert (ITE) is designed, which incorporates grouped self-attention operations to perform short-range sequence modeling within token groups at each scale. Meanwhile, a cross-scale Mamba expert (CME) is introduced, which integrates a cross-scale serialization mechanism and bidirectional Mamba block for long-range sequence modeling, further exploring the interactions and complementarity between token groups across different scales. Several hybrid strategies for integrating the ITE and CME are investigated to maximize their complementarity, including parallel, interval, and serial structures. Extensive experiments demonstrate that the proposed HSI-MFormer significantly outperforms the state-of-the-art Transformer-based and Mamba-based HSI classification methods. © 1980-2012 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Mortar: Morphing the Bit Level Sparsity for General Purpose Deep Learning Acceleration

Mortar: Morphing the Bit Level Sparsity for General Purpose ...

引用

Asia and South Pacific Design Automation Conference

作者： Yunhung Gao Hongyan Li Kevin Zhang Xueru Yu Hang Lu School of Electronics Engineering and Computer Science Peking University Beijing China State Key Lab of Computer Architecture Institute of Computing Technology CAS University of Chinese Academy of Sciences Beijing China Shanghai Integrated Circuits R&D Center Co. Ltd Shanghai China

Vanilla Deep Neural Networks (DNN) after training are represented with native floating-point 32 (fp32) weights. We observe that the bit-level sparsity of these weights is very abundant in the mantissa and can be directly exploited to speed up model inference. In this paper, we propose Mortar, an off-line/on-line collaborated approach for fp32 DNN acceleration, which includes two parts: first, an off-line bit sparsification algorithm to construct the target formulation by “mantissa morphing”, which maintains higher model accuracy while increasing bit-level sparsity; second, the associating hardware accelerator architecture to speed up the on-line fp32 inference through manipulating the enlarged bit sparsity. We highlight the following results by evaluating various deep learning tasks, including image classification, object detection, video understanding, video & image super-resolution, etc.: We (1) increase bit-level sparsity up to 1.28~2.51x with only a negligible -0.09~0.23% accuracy loss, (2) maintain on average 3.55% higher model accuracy while increasing more bit-level sparsity than the baseline, (3)and our hardware accelerator outperforms up to 4.8x over the baseline, with an area of 0.031 mm 2 and power of 68.58mW.

关键词： Deep learning Training Mortar Superresolution Software algorithms Inference algorithms Software

来源：评论

学校读者我要写书评

暂无评论

Fast Maxwell Solvers Based on Exact Discrete Eigen-Decompositions I. Two-Dimensional Case

arXiv

引用

arXiv 2025年

作者： Wang, Lixiu Jia, Lueling Cao, Zijian Li, Huiyuan Zhang, Zhimin School of Mathematics and Physics University of Science and Technology Beijing Beijing100083 China School of Mathematics and Statistics Shandong Normal University Jinan250358 China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing101408 China State Key Laboratory of Computer Science Laboratory of Parallel Computing Institute of Software Chinese Academy of Sciences Beijing100190 China Department of Mathematics Wayne State University DetroitMI48202 United States

In this paper, we propose fast solvers for Maxwell’s equations in rectangular domains. We first discretize the simplified Maxwell’s eigenvalue problems by employing the lowest-order rectangular Nédélec elements and derive the discrete eigen-solutions explicitly, providing a Hodge-Helmholtz decomposition framework at the discrete level. Based on exact eigen-decompositions, we further design fast solvers for various Maxwell’s source problems, guaranteeing either the divergence-free constraint or the Gauss’s law at the discrete level. With the help of fast sine/cosine transforms, the computational time grows asymptotically as O(n2 log n) with n being the number of grids in each direction. Our fast Maxwell solvers outperform other existing Maxwell solvers in the literature and fully rival fast scalar Poisson/Helmholtz solvers based on trigonometric transforms in either efficiency, robustness, or storage complexity. It is also utilized to perform an efficient pre-conditioning for solving Maxwell’s source problems with variable coefficients. Finally, numerical experiments are carried out to illustrate the effectiveness and efficiency of the proposed fast *** Codes 65N30, 35Q61, 65F05, 68W40 © 2025, CC BY.

关键词： Maxwell equations

来源：评论

学校读者我要写书评

暂无评论

Optimal Bilinear Equalizer for Cell-Free Massive MIMO Systems over Correlated Rician Channels

引用

IEEE Transactions on Signal Processing 2025年

作者： Wang, Zhe Zhang, Jiayi Bjornson, Emil Niyato, Dusit Ai, Bo Beijing Jiaotong University State Key Laboratory of Advanced Rail Autonomous Operation Beijing100044 China Beijing Jiaotong University School of Electronics and Information Engineering Beijing100044 China KTH Royal Institute of Technology Department of Computer Science Stockholm114 28 Sweden Nanyang Technological University College of Computing & Data Science 639798 Singapore

In this paper, we explore the low-complexity optimal bilinear equalizer (OBE) combining scheme design for cell-free massive multiple-input multiple-output networks with spatially correlated Rician fading *** provide a spectral efficiency (SE) performance analysis framework for both the centralized and distributed processing schemes with bilinear equalizer (BE)-structure combining schemes applied. The BE-structured combining is a set of schemes that are constructed by the multiplications of channel statistics-based BE matrices and instantaneous channel estimates. Notably, we derive closed-form achievable SE expressions for centralized and distributed BE-structured combining schemes. We propose one centralized and two distributed OBE schemes: Centralized OBE (C-OBE), Distributed OBE based on Global channel statistics (DG-OBE), and Distributed OBE based on Local channel statistics (DL-OBE), which maximize their respective SE expressions. OBE matrices in these schemes are tailored based on varying levels of channel statistics. Notably, we obtain new and insightful closed-form results for the C-OBE, DG-OBE, and DL-OBE combining schemes. Numerical results demonstrate that the proposed OBE schemes can achieve excellent SE, even in scenarios with severe pilot contamination. © 2025 IEEE. All rights reserved.

关键词： Fading channels

来源：评论

学校读者我要写书评

暂无评论

Focus, Distinguish, and Prompt: Unleashing CLIP for Efficient and Flexible Scene Text Retrieval

arXiv

引用

arXiv 2024年

作者： Zeng, Gangyan Zhang, Yuan Wei, Jin Yang, Dongbao Zhang, Peng Gao, Yiwen Qin, Xugong Zhou, Yu School of Cyber Science and Engineering Nanjing University of Science and Technology China State Key Laboratory of Media Convergence and Communication Communication University of China China Institute of Information Engineering Chinese Academy of Sciences China Laboratory for Advanced Computing and Intelligence Engineering China TMCC College of Computer Science Nankai University China

Scene text retrieval aims to find all images containing the query text from an image gallery. Current efforts tend to adopt an Optical Character Recognition (OCR) pipeline, which requires complicated text detection and/or recognition processes, resulting in inefficient and inflexible retrieval. Different from them, in this work we propose to explore the intrinsic potential of Contrastive Language-Image Pre-training (CLIP) for OCR-free scene text retrieval. Through empirical analysis, we observe that the main challenges of CLIP as a text retriever are: 1) limited text perceptual scale, and 2) entangled visual-semantic concepts. To this end, a novel model termed FDP (Focus, Distinguish, and Prompt) is developed. FDP first focuses on scene text via shifting the attention to the text area and probing the hidden text knowledge, and then divides the query text into content word and function word for processing, in which a semantic-aware prompting scheme and a distracted queries assistance module are utilized. Extensive experiments show that FDP significantly enhances the inference speed while achieving better or competitive retrieval accuracy compared to existing methods. Notably, on the IIIT-STR benchmark, FDP surpasses the state-of-the-art model by 4.37% with a 4 times faster speed. Furthermore, additional experiments under phrase-level and attribute-aware scene text retrieval settings validate FDP’s particular advantages in handling diverse forms of query text. The source code will be publicly available at https://***/Gyann-z/FDP. © 2024, CC BY-NC-SA.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A method of generating random bits by using electronic bipolar memristor

引用

Chinese Physics B 2020年第4期29卷 535-540页

作者： Bin-Bin Yang Nuo Xu Er-Rui Zhou Zhi-Wei Li Cheng Li Pin-Yun Yi Liang Fang Institute for Quantum Information&State Key Laboratory of High Performance Computing College of ComputerNational University of Defense TechnologyChangsha 410073China College of Computer National University of Defense TechnologyChangsha 410073China College of Electronic Science and Engineering National University of Defense TechnologyChangsha 410073China

The intrinsic stochasticity of resistance switching process is one of the holdblocks for using memristor as a fundamental element in the next-generation nonvolatile ***,such a weakness can be used as an asset for generating the random bits,which is valuable in a hardware security *** this work,a forming-free electronic bipolar Pt/Ti/Ta2O5/Pt memristor is successfully fabricated to investigate the merits of generating random bits in such a *** resistance switching mechanism of the fabricated device is ascribed to the electric field conducted electrons trapping/de-trapping in the deep-energy-level traps produced by the"oxygen grabbing"*** stochasticity of the electrons trapping/detrapping governs the random distribution of the set/reset switching voltages of the device,making a single memristor act as a random bit in which the resistance of the device represents information and the applied voltage pulse serves as the triggering *** physical implementation of such a random process provides a method of generating the random bits based on memristors in hardware security applications.

关键词： memristor resistance switching electrons trapping/de-trapping random bits

来源：评论

学校读者我要写书评

暂无评论

Streamline Ring ORAM Accesses through Spatial and Temporal Optimization

Streamline Ring ORAM Accesses through Spatial and Temporal O...

引用

IEEE Symposium on High-Performance computer architecture

作者： Dingyuan Cao Mingzhe Zhang Hang Lu Xiaochun Ye Dongrui Fan Yuezhi Che Rujia Wang Tsinghua University Beijing China State Key Laboratory of Computer Architecture ICT CAS Beijing China State Key Laboratory of Mathematical Engineering and Advanced Computing China Illinois Institute of Technology

Memory access patterns could leak temporal and spatial information in a sensitive program; therefore, obfuscated memory access patterns are desired from the security perspective. Oblivious RAM (ORAM) has been the favored candidate to eliminate the access pattern leakage through randomly remapping data blocks around the physical memory space. Meanwhile, accessing memory with ORAM protocols results in significant memory bandwidth overhead. For each memory request, after going through the ORAM obfuscation, the main memory needs to service tens of actual memory accesses, and only one real access out of them is useful for the program execution. Besides, to ensure the memory bus access patterns are indistinguishable, extra dummy blocks need to be stored and transmitted, which cause memory space waste and poor performance. In this work, we introduce a new framework, String ORAM, that accelerates the Ring ORAM accesses with Spatial and Temporal optimization schemes. First, we identify that dummy blocks could significantly waste memory space and propose a compact ORAM organization that leverages the real blocks in memory to obfuscate the memory access pattern. Then, we identify the inefficiency of current transaction-based Ring ORAM scheduling on DRAM devices and propose an effective scheduling technique that can overlap the time spent on row buffer misses while ensuring correctness and security. With a minimal modification on the hardware and software, and negligible impact on security, the framework reduces 30.05% execution time and up to 40% memory space overhead compared to the state-of-the-art bandwidth-efficient Ring ORAM.

关键词： Protocols Space debris Memory management Random access memory Performance gain Scheduling Software

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Recognition of Unknown Objects for Open-World Object Detection

arXiv

引用

arXiv 2023年

作者： Fang, Ruohuan Pang, Guansong Zhou, Lei Bai, Xiao Zheng, Jin School of Computer Science and Engineering State Key Laboratory of Software Development Environment Jiangxi Research Institute Beihang University China School of Computing and Information Systems Singapore Management University Singapore School of Computer Science and Artificial Intelligence Wuhan University of Technology China

Open-World Object Detection (OWOD) extends object detection problem to a realistic and dynamic scenario, where a detection model is required to be capable of detecting both known and unknown objects and incrementally learning newly introduced knowledge. Current OWOD models, such as ORE and OW-DETR, focus on pseudo-labeling regions with high objectness scores as unknowns, whose performance relies heavily on the supervision of known objects. While they can detect the unknowns that exhibit similar features to the known objects, they suffer from a severe label bias problem that they tend to detect all regions (including unknown object regions) that are dissimilar to the known objects as part of the background. To eliminate the label bias, this paper proposes a novel approach that learns an unsupervised discriminative model to recognize true unknown objects from raw pseudo labels generated by unsupervised region proposal methods. The resulting model can be further refined by a classification-free self-training method which iteratively extends pseudo unknown objects to the unlabeled regions. Experimental results show that our method 1) significantly outperforms the prior SOTA in detecting unknown objects while maintaining competitive performance of detecting known object classes on the MS COCO dataset, and 2) achieves better generalization ability on the LVIS and Objects365 datasets. Code is available at https://***/frh23333/mepu-owod. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

VADER: Leveraging the Natural Variation of Hardware to Enhance Adversarial Attack

VADER: Leveraging the Natural Variation of Hardware to Enhan...

引用

Asia and South Pacific Design Automation Conference

作者： Hao Lv Bing Li Ying Wang Cheng Liu Lei Zhang Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Capital Normal University State Key Laboratory of Computer Architecture

ISBN: (数字)9781450379991

ISBN: (纸本)9781728180571

Adversarial attacks have been viewed as the primary threat to the security of neural networks. Hence, extensive adversarial defense techniques have been proposed to protect the neural networks from adversarial attacks, allowing for the application of neural networks to the security-sensitive tasks. Recently, the emerging devices, e.g., Resistive RAM (RRAM), attracted extensive attention for establishing the hardware platform for neural networks to tackle the inadequate computing capability of the traditional computing platform. Though the emerging devices exhibit the instinct instability issues due to the advanced manufacture technology, including hardware variations and defects, the error-resilience capability of neural networks enables the wide deployment of neural networks on the emerging devices. In this work, we find that the natural instability in emerging devices impairs the security of neural networks. Specifically, we design an enhanced adversarial attack, Variation-oriented ADvERsarial (VADER) attack which leverages the inherent hardware variations in RRAM chips to penetrate the protection of adversarial defenses and mislead the prediction of neural networks. We evaluated the effectiveness of VADER across various protected neural network models and the result shows that VADER achieves higher success attack rate over other adversarial attacks.

关键词： Semiconductor lasers Neural networks Resistive RAM Hardware Security Laser beams Task analysis

来源：评论

学校读者我要写书评

暂无评论

CNN+LSTM Accelerated Turbulent Flow Simulation with Link-Wise Artificial Compressibility Method 21

CNN+LSTM Accelerated Turbulent Flow Simulation with Link-Wis...

引用

50th International Conference on Parallel Processing, ICPP 2021

作者： Fan, Sijiang Fei, Jiawei Guo, Xiao-Wei Yang, Canqun Revell, Alistair Institute for Quantum Information and State Key Laboratory of High Performance Computing College of Computer National University of Defense Technology Changsha China School of Mechanical Aerospace and Civil Engineering University of Manchester Manchester United Kingdom

ISBN: (纸本)9781450390682

The simulation of turbulent flow, the most common form of fluid, is indispensable in computational fluid dynamics (CFD). The synthetic eddy method (SEM) generates the turbulent inflow and is adopted as the inlet boundary condition of simulation. However, SEM is time-consuming and can significantly slow down the simulation process which occupies 58% of the whole computational time. This is highly inefficient especially since SEM is only used as the inlet. In this paper we propose an efficient alternative. In particular, we leverage CNN+LSTM to replace SEM to obtain the turbulence statistics and combine it with link-wise artificial compressibility method (LW-ACM), which is a fast numerical method of CFD. We validate the predicted results by CNN+LSTM and prove that our model can provide the correct turbulence statistics even after a long time. Experiment results show that our CNN+LSTM module achieves over 15 × speedup compared with SEM, which greatly reduces the time consumption of turbulent inflow generation (from 58% to 7%). As a result, the whole time of turbulent flow simulation is more than halved. Compared with a newly released GPU-accelerated standard lattice Boltzmann method solver, our combination of CNN+LSTM and LW-ACM is about 8.6 × faster. Among all studies reported to date, our work is the fastest implementation for simulating turbulent channel flow, an important step for the field of fast CFD analysis. © 2021 ACM.

关键词： Computational fluid dynamics

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：