检索结果-内蒙古大学图书馆

Generating Image Counterfactuals in Deep Learning Models Without the Aid of Generative Models

IEEE signal processing LETTERS 2025年 32卷 1495-1499页

作者： Xu, Ao Li, Zihao Zhang, Yukai Wu, Tieru Jilin Univ Sch Artificial Intelligence Changchun 130000 Peoples R China

With the rapid development of artificial intelligence, particularly the rise of deep learning, the importance of Explainable Artificial Intelligence has become increasingly prominent. Among its key techniques, counterfactual explanation plays a crucial role in understanding the decision-making mechanisms of opaque models. However, the high dimensionality and complex feature patterns of image data pose significant challenges for the task of generating counterfactuals for images. Existing literature has proposed various algorithms based on different assumptions, many of which rely on the existence of appropriate generative models. Some of these assumptions, particularly the assumption regarding the existence of generative models, may be overly stringent. To address this issue, this letter introduces a novel assumption-free image counterfactual generation algorithm, DFO-S, based on Score Matching and gradient-free optimization techniques. The proposed method achieves high-quality counterfactual generation without relying on generative models. Through extensive empirical analysis, we demonstrate the significant superiority of our method in terms of performance.

关键词： signal processing algorithms Artificial intelligence Closed box Deep learning Vectors Optimization Data models Computational modeling Training Smoothing methods explainable artificial intelli- gence image counterfactual explanation

来源：评论

学校读者我要写书评

暂无评论

Optimizing the Order of Modes in Tensor Train Decomposition

引用

IEEE signal processing LETTERS 2025年 32卷 1361-1365页

作者： Tichavsky, Petr Straka, Ondrej Czech Acad Sci Inst Informat Theory & Automat Prague 18200 Czech Republic Univ West Bohemia Dept Cybernet Bohemia 30614 Czech Republic

The tensor train (TT) is a popular way of representing high-dimensional hyper-rectangular data structures called tensors. It is widely used, for example, in quantum chemistry under the name "matrix product state". The complexity of the TT model mainly depends on the bond dimensions that connect TT cores, constituting the model. Unlike canonical polyadic decomposition, the TT model complexity may depend on the order of the modes/indices in the data structures or the order of the core tensors in the TT, in general. This letter aims to provide methods for optimizing the order of the modes to reduce the bond dimensions. Since the number of possible orderings of the cores is exponentially high, we propose a greedy algorithm that provides a suboptimal solution. We consider three problem setups, i.e., specifications of the tensor: tensor given by a list of all its elements, tensor described by a TT model with some default order of the modes, and tensor obtained by sampling a multivariate function.

关键词： Tensors signal processing algorithms Numerical models Computational modeling Matrix decomposition Complexity theory Optimization Quantum chemistry Chemical elements Upper bound tensor train decomposition matrix product state

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Subaperture Preprocessing for Multireceiver Wide-Beam SAS Imaging

引用

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 2025年 22卷

作者： Zhang, Jiafeng Cheng, Guangli Tang, Jinsong Wu, Haoran Naval Univ Engn Inst Elect Engn Wuhan 430000 Peoples R China

The multireceiver synthetic aperture sonar (SAS) data are usually converted to equivalent monostatic data through the displaced phase center approximation (DPCA) before the monostatic imaging. However, the DPCA error is azimuth-variant in the wide-beam case, resulting in the traditional algorithms compensating for the DPCA error in the extended Doppler domain of each receiver, which obviously increases the computational complexity. To solve the problem, this letter analyzes the space-variant characteristics of the DPCA error and discovers that the DPCA error exhibits significant receiver-variant and weak azimuth-variant characteristics. Based on this, a subaperture preprocessing is proposed to reduce computational complexity without sacrificing imaging accuracy. The proposed algorithm compensates for the DPCA error uniformly within the subaperture using the DPCA error at the center of the subaperture and then superposes all subapertures coherently to obtain equivalent monostatic data. The weak azimuth-variant characteristic ensures that the subapertures are very sparse, Additionally, the algorithm allows data recording and preprocessing to be synchronized, significantly reducing the imaging waiting time further. The simulated data and actual data experiments verify the effectiveness of the proposed algorithm.

关键词： Receivers Synthetic aperture sonar signal processing algorithms Imaging Doppler effect Azimuth Approximation algorithms Apertures Transmitters Geoscience and remote sensing Azimuth-variant characteristic displaced phase center approximation (DPCA) preprocessing real time subaperture synthetic aperture sonar (SAS) wide-beam

来源：评论

学校读者我要写书评

暂无评论

The AAV Tracking: Fusion and Adversarial of Global and Local Information

引用

IEEE signal processing LETTERS 2025年 32卷 96-100页

作者： Zhao, Yuhui Yang, Ruifeng Guo, Chenxia Hao, Yongchang North Univ China Sch Instrument & Elect Taiyuan 030051 Peoples R China

The widespread application of autonomous aerial vehicles (AAVs) in various fields has driven the development of target tracking based on AAV perspectives. Among the existing tracking algorithms, tracking by detection performs well, which depends heavily on the robustness of the object detection and re-identification algorithms. The images obtained from the perspective of drones have complex backgrounds and small image sizes, making tracking tasks more difficult. We introduce the parallel space and channel attention module to optimize the object detection algorithm, exploring compelling spatial and channel dimensions features. We propose the channel-spatial fully connected information extraction module, which explores the connections between pixels from a spatial perspective and considers the relationships between feature maps from a channel perspective, compensating for the shortcomings of convolutional networks in extracting global features. To obtain more discriminative feature descriptions, we design the global-local-residual feature fusion module that divides the image into global and local parts and gets the difference between local information and other information, which helps with data correlation and label matching. Experiments show that our method improves the mAP by 0.4% and AP by 2.1% in the object detection task. In the re-identification task, top-1 accuracy is improved by 5.3% and mAP by 11.3%.

关键词： Target tracking Accuracy signal processing algorithms Object detection Feature extraction Information retrieval Autonomous aerial vehicles Robustness Convolutional neural networks Drones re-identification tracking-by-detection autonomous aerial vehicle (AAV) tracking

来源：评论

学校读者我要写书评

暂无评论

Outlier-Robust Underwater Navigation Using a Dual-Robust-Kernel Kalman Filter

引用

IEEE signal processing LETTERS 2025年 32卷 1485-1489页

作者： Bao, Jingqiang Mu, Xiaokai Yu, Xiang Zhu, Zhongben Qin, Hongde Harbin Engn Univ Nanhai Inst Sanya 572024 Peoples R China Harbin Engn Univ Qingdao Innovat & Dev Base Qingdao 266000 Peoples R China China Univ Petr East China Coll Mech & Elect Engn Qingdao 266580 Peoples R China Harbin Engn Univ Natl Key Lab Autonomous Marine Vehicle Technol Harbin 15001 Peoples R China

This paper addresses the challenge of robust navigation for unmanned underwater vehicles(UUVs) operating in harsh environments, where sensor data is often contaminated by outliers. To mitigate this issue, a dual-robust-kernel Kalman filter is proposed, which sequentially applies the Huber and Agarwal kernels during the measurement update phase. This method effectively combines the exceptional numerical stability of the Huber kernel with the superior robustness of the Agarwal kernel, leveraging their complementary strengths. The UUV experiments validate the proposed algorithm by benchmarking it against several representative methods.

关键词： Kernel Kalman filters signal processing algorithms Robustness Covariance matrices Vectors Numerical stability Information filters Training Pollution measurement Kalman filter robust filtering Kernel function

来源：评论

学校读者我要写书评

暂无评论

Outlier Indicator Based Projection Fuzzy K-Means Clustering for Hyperspectral Image

引用

IEEE signal processing LETTERS 2025年 32卷 496-500页

作者： Liu, Xinze Yang, Xiaojun Zhang, Jiale Wang, Jing Nie, Feiping Guangdong Univ Technol Sch Informat Engn Guangzhou 510006 Peoples R China Guangdong Univ Technol Guangdong Prov Key Lab Informat Photon Technol China Key Lab Photon Technol Integrated Sensing & Minist Educ China Guangzhou 510006 Peoples R China Northwestern Polytech Univ Sch Comp Sci Xian 710072 Peoples R China Northwestern Polytech Univ Ctr OPT IMagery Anal & Learning OPTIMAL Xian 710072 Peoples R China

The application of hyperspectral image (HSI) clustering has become widely used in the field of remote sensing. Traditional fuzzy K-means clustering methods often struggle with HSI data due to the significant levels of noise, consequently resulting in segmentation inaccuracies. To address this limitation, this letter introduces an innovative outlier indicator-based projection fuzzy K-means clustering (OIPFK) algorithm for clustering of HSI data, enhancing the efficacy and robustness of previous fuzzy K-means methodologies through a two-pronged strategy. Initially, an outlier indicator vector is constructed to identify noise and outliers by computing the distances between each data point in a reduced dimensional space. Subsequently, the OIPFK algorithm incorporates the fuzzy membership relationships between samples and clustering centers within this lower-dimensional framework, along with the integration of the outlier indicator vectors, to significantly mitigates the influence of noise and extraneous features. Moreover, an efficient iterative optimization algorithm is employed to address the optimization challenges inherent to OIPKM. Experimental results from three real-world hyperspectral image datasets demonstrate the effectiveness and superiority of our proposed method.

关键词： Hyperspectral imaging Optimization Noise Vectors Clustering algorithms Robustness Linear programming signal processing algorithms Photonics Indexes Projection fuzzy K-means hyperspectral image robustness outliers clustering

来源：评论

学校读者我要写书评

暂无评论

Inpainting-Driven Graph Learning via Explainable Neural Networks

引用

IEEE signal processing LETTERS 2025年 32卷 111-115页

作者： Batreddy, Subbareddy Mishra, Pushkal Kakarla, Yaswanth Siripuram, Aditya Indian Inst Technol Hyderabad Hyderabad 502284 India

Given partial measurements of a time-varying graph signal, we propose an algorithm to simultaneously estimate both the underlying graph topology and the missing measurements. The proposed algorithm operates by training an interpretable neural network, designed from the unrolling framework. The proposed technique can be used as a graph learning and/or a graph signal reconstruction algorithm. This work builds on prior work in graph learning by tailoring the learned graph to the signal reconstruction task;and also enhances prior work in graph signal reconstruction by allowing the underlying graph to be unknown.

关键词： signal processing algorithms Neural networks Optimization Laplace equations Data models Covariance matrices Biological neural networks Training Minimization Topology Graph temporal data graph signal processing graph topology inference unrolling

来源：评论

学校读者我要写书评

暂无评论

Adaptive Superpixel-Guided Non-Homogeneous Image Dehazing

引用

IEEE signal processing LETTERS 2025年 32卷 591-595页

作者： Zhang, Hao Lu, Ping Qi, Te Xu, Yan Zeng, Tieyong Chinese Univ Hong Kong Dept Math Hong Kong Peoples R China ZTE Corp Shenzhen 518057 Peoples R China

Image dehazing is regarded as a fundamental image processing task with a major impact on higher-level imaging tasks. Many existing haze removal methods are designed for homogeneous haze, but in real-world cases, the haze is normally non-homogeneous. Superpixels, which segment an image into a set of closely spaced regions, can be employed in real-world scenarios to deal with non-homogeneous haze. In our paper, an adaptive non-homogeneous image dehazing approach that utilizes the superpixel-guided algorithm is designed to segment different hazy regions. Considering that both ambient light and transmission map estimation have a significant impact on the results, our research focuses on the development of a variational dehazing model that takes into account non-uniform ambient light and non-uniform transmission maps to address varying levels of haze. A series of numerical results illustrate the superiority and efficacy of our method.

关键词： Scattering signal processing algorithms Mathematical models Atmospheric modeling Image segmentation Optimization Feature extraction Numerical models Lighting Estimation Image dehazing non-homogeneous superpixel algorithm

来源：评论

学校读者我要写书评

暂无评论

MPHR: A Robust Algorithm for Estimating Nodal Head in Water Networks

引用

IEEE signal processing LETTERS 2025年 32卷 856-860页

作者： Maggang, Amin A. Tay, David B. Gong, Jinzhe Josey, Brendan M. Deakin Univ Sch IT Geelong Vic 3220 Australia Deakin Univ Sch Engn Geelong Vic 3220 Australia

The monitoring of the pressure, in terms of nodal head values, is an important task in water distribution networks, to ascertain the integrity of the components, e.g. pipes. The placement of sensors at desired locations, in realistic situations, is a challenging problem due to various practical constraints. In practice, nodal heads are estimated at locations with no sensors. In this work, we propose an algorithm, termed mean percolation head reconstruction (MPHR), for estimating the nodal heads. The iterative algorithm is inspired by concepts from graph signal processing (GSP), and consists of two stages. The first stage is equivalent to the percolation of known values across the network, and the second stage performs some form of constrained consensus computations. The algorithm is robust against uncertainty in the network parameters and suboptimal placement of sensors. Extensive simulations have demonstrated the superiority of MPHR compared to other proposed reconstruction algorithms.

关键词： signal processing algorithms Sensors Magnetic heads Pressure sensors Head Distribution networks Vectors Uncertainty Training signal processing Graph signal processing (GSP) graph signal reconstruction nodal head estimation water distribution network

来源：评论

学校读者我要写书评

暂无评论

Windowed Quantum Phase Estimation: signal processing Approach to a Quantum Algorithm

Windowed Quantum Phase Estimation: Signal Processing Approac...

引用

International Conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Xue Wen Samsung Research China Beijing

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Quantum phase estimation (QPE) is a key ingredient to some of the most important results of quantum computing. In this paper we examine the QPE algorithm through the looking glass of signal processing. This perspective helps to reveal a relation between the distribution of quantum phase estimates and the shape of the discrete Fourier transform (DFT) of a pure complex sinusoid. Guided by spectral shaping theory, we design a class of quantum circuits that are the quantum equivalents to classical windowing operators, for a family of window functions that have compact support in the DFT. These circuits have complexity linear in the number of target qubits, strictly lower than QPE itself. Using these circuits we are able to reshape QPE's performance characteristics. In particular, the asymptotic marginal cost for 2x reduction in fail rate (2x improvement in reliability) is reduced from one qubit to 1/3 qubit or lower, which can help reduce the qubit budget for a preset accuracy-reliability target, especially in high-reliability scenarios. This makes a demonstration of the value of signal processing principles in designing quantum algorithms.

关键词： Quantum algorithm Shape Qubit Discrete Fourier transforms Phase estimation signal processing algorithms Quantum mechanics signal processing Windows Speech processing

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：