检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Zhu, Zijian Zhang, Yichi Chen, Hai Dong, Yinpeng Zhao, Shu Ding, Wenbo Zhong, Jiachen Zheng, Shibao Institute of Image Communication and Network Engineering Shanghai Jiao Tong University China Dept. of Comp. Sci. and Tech. Institute for AI THBI Lab BNRist Center Tsinghua University China Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education School of Computer Science and Technology Anhui University Information Materials and Intelligent Sensing Laboratory of Anhui Province China SAIC Motor AI Lab China Zhongguancun Laboratory China

3D object detection is an essential perception task in autonomous driving to understand the environments. The Bird’s-Eye-View (BEV) representations have significantly improved the performance of 3D detectors with camera inputs on popular benchmarks. However, there still lacks a systematic understanding of the robustness of these vision-dependent BEV models, which is closely related to the safety of autonomous driving systems. In this paper, we evaluate the natural and adversarial robustness of various representative models under extensive settings, to fully understand their behaviors influenced by explicit BEV features compared with those without BEV. In addition to the classic settings, we propose a 3D consistent patch attack by applying adversarial patches in the 3D space to guarantee the spatiotemporal consistency, which is more realistic for the scenario of autonomous driving. With substantial experiments, we draw several findings: 1) BEV models tend to be more stable than previous methods under different natural conditions and common corruptions due to the expressive spatial representations;2) BEV models are more vulnerable to adversarial noises, mainly caused by the redundant BEV features;3) Camera-LiDAR fusion models have superior performance under different settings with multi-modal inputs, but BEV fusion model is still vulnerable to adversarial noises of both point cloud and image. These findings alert the safety issue in the applications of BEV detectors and could facilitate the development of more robust models. Code available at https: //***/zzj403/BEV_Robust. Copyright © 2023, The Authors. All rights reserved.

关键词： Birds

来源：评论

学校读者我要写书评

暂无评论

6G Wireless Communications: From Far-field Beam Steering to Near-field Beam Focusing

arXiv

引用

arXiv 2022年

作者： Zhang, Haiyang Shlezinger, Nir Guidi, Francesco Dardari, Davide Eldar, Yonina C. The School of Communication and Information Engineering Nanjing University of Posts and Telecommunications China The School of Electrical and Computer Engineering Ben-Gurion University Israel Electronics Telecommunications and Information Technologies from Ecole Polytechnique Paris-Tech France Computer Science Specialty The University of Bologna Italy IEIIT-CNR Italy The University of Bologna WiLAB-CNIT Italy The Department of Math and Computer Science Weizmann Institute of Science Israel The Center for Biomedical Engineering and Signal Processing The Israel Academy of Sciences and Humanities IEEE Fellow EURASIP Fellow

6G networks will be required to support higher data rates, improved energy efficiency, lower latency, and more diverse users compared with 5G systems. To meet these requirements, electrically extremely large-scale antenna arrays are envisioned to be key physical-layer technologies. As a consequence, it is expected that some portion of future 6G wireless communications may take place in the radiating near-field (Fresnel) region, in addition to the far-field operation as in current wireless technologies. In this article, we discuss the opportunities and challenges that arise in radiating near-field communications. We begin by discussing the key physical characteristics of near-field communications, where the standard plane-wave propagation assumption no longer holds, and clarifying its implication on the modelling of wireless channels. Then, we elaborate on the ability to leverage spherical wavefronts via beam focusing, highlighting its advantages for 6G systems. We point out several appealing application scenarios which, with proper design, can benefit from near-field operation, including interference mitigation in multi-user communications, accurate localization and focused sensing, as well as wireless power transfer with minimal energy pollution. We conclude by discussing some of the design challenges and research directions that are yet to be explored to fully harness the potential of near-field operation. Copyright © 2022, The Authors. All rights reserved.

关键词： Antenna arrays

来源：评论

学校读者我要写书评

暂无评论

Gradients as a Measure of Uncertainty in Neural Networks

Gradients as a Measure of Uncertainty in Neural Networks

引用

IEEE International Conference on image processing

作者： Jinsol Lee Ghassan AlRegib OLIVES at the Center for Signal and Information Processing School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

Despite tremendous success of modern neural networks, they are known to be overconfident even when the model encounters inputs with unfamiliar conditions. Detecting such inputs is vital to preventing models from making naive predictions that may jeopardize real-world applications of neural networks. In this paper, we address the challenging problem of devising a simple yet effective measure of uncertainty in deep neural networks. Specifically, we propose to utilize backpropagated gradients to quantify the uncertainty of trained models. Gradients depict the required amount of change for a model to properly represent given inputs, thus providing a valuable insight into how familiar and certain the model is regarding the inputs. We demonstrate the effectiveness of gradients as a measure of model uncertainty in applications of detecting unfamiliar inputs, including out-of-distribution and corrupted samples. We show that our gradient-based method outperforms state-of-the-art methods by up to 4.8% of AUROC score in out-of-distribution detection and 35.7% in corrupted input detection.

关键词： Uncertainty Training Measurement uncertainty Computational modeling Neural networks Data models Predictive models

来源：评论

学校读者我要写书评

暂无评论

Contrastive Explanations In Neural Networks

Contrastive Explanations In Neural Networks

引用

IEEE International Conference on image processing

作者： Mohit Prabhushankar Gukyeong Kwon Dogancan Temel Ghassan AlRegib OLIVES at the Center for Signal and Information Processing School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

Visual explanations are logical arguments based on visual features that justify the predictions made by neural networks. Current modes of visual explanations answer questions of the form `Why P.7'. These Why questions operate under broad contexts thereby providing answers that are irrelevant in some cases. We propose to constrain these Why questions based on some context Q so that our explanations answer contrastive questions of the form `Why P, rather than Q.7'. In this paper, we formalize the structure of contrastive visual explanations for neural networks. We define contrast based on neural networks and propose a methodology to extract defined contrasts. We then use the extracted contrasts as a plug-in on top of existing `Why P?' techniques, specifically Grad-CAM. We demonstrate their value in analyzing both networks and data in applications of large-scale recognition, fine-grained recognition, subsurface seismic analysis, and image quality assessment.

关键词： Visualization Neural networks Manifolds image recognition image quality Automobiles image color analysis

来源：评论

学校读者我要写书评

暂无评论

Novelty Detection Through Model-Based Characterization of Neural Networks

Novelty Detection Through Model-Based Characterization of Ne...

引用

IEEE International Conference on image processing

作者： Gukyeong Kwon Mohit Prabhushankar Dogancan Temel Ghassan AlRegib OLIVES at the Center for Signal and Information Processing School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

In this paper, we propose a model-based characterization of neural networks to detect novel input types and conditions. Novelty detection is crucial to identify abnormal inputs that can significantly degrade the performance of machine learning algorithms. Majority of existing studies have focused on activation-based representations to detect abnormal inputs, which limits the characterization of abnormality from a data perspective. However, a model perspective can also be informative in terms of the novelties and abnormalities. To articulate the significance of the model perspective in novelty detection, we utilize back-propagated gradients. We conduct a comprehensive analysis to compare the representation capability of gradients with that of activation and show that the gradients outperform the activation in novel class and condition detection. We validate our approach using four image recognition datasets including MNIST, Fashion-MNIST, CIFAR10, and CURE-TSR. We achieve a significant improvement on all four datasets with an average AUROC of 0.953, 0.918, 0.582, and 0.746, respectively.

关键词： image reconstruction Neural networks Training Feature extraction Loss measurement Backpropagation Decoding

来源：评论

学校读者我要写书评

暂无评论

Implicit Saliency In Deep Neural Networks

Implicit Saliency In Deep Neural Networks

引用

IEEE International Conference on image processing

作者： Yutong Sun Mohit Prabhushankar Ghassan AlRegib OLIVES at the Center for Signal and Information Processing School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

In this paper, we show that existing recognition and localization deep architectures, that have not been exposed to eye tracking data or any saliency datasets, are capable of predicting the human visual saliency. We term this as implicit saliency in deep neural networks. We calculate this implicit saliency using expectancy-mismatch hypothesis in an unsupervised fashion. Our experiments show that extracting saliency in this fashion provides comparable performance when measured against the state-of-art supervised algorithms. Additionally, the robustness outperforms those algorithms when we add large noise to the input images. Also, we show that semantic features contribute more than low-level features for human visual saliency detection. Based on these properties and performances, our proposed method greatly lowers the threshold for saliency detection in terms of required data and bridges the gap between human visual saliency and model saliency.

关键词： Feature extraction Visualization Semantics Saliency detection Convolution Robustness Neural networks

来源：评论

学校读者我要写书评

暂无评论

On the Structures of Representation for the Robustness of Semantic Segmentation to Input Corruption

On the Structures of Representation for the Robustness of Se...

引用

IEEE International Conference on image processing

作者： Charles Lehman Dogancan Temel Ghassan Alregib OLIVES at the Center for Signal and Information Processing School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA USA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

Semantic segmentation is a scene understanding task at the heart of safety-critical applications where robustness to corrupted inputs is essential. Implicit Background Estimation (IBE) has demonstrated to be a promising technique to improve the robustness to out-of-distribution inputs for semantic segmentation models for little to no cost. In this paper, we provide analysis comparing the structures learned as a result of optimization objectives that use Softmax, IBE, and Sigmoid in order to improve understanding their relationship to robustness. As a result of this analysis, we propose combining Sigmoid with IBE (SCrIBE) to improve robustness. Finally, we demonstrate that SCrIBE exhibits superior segmentation performance aggregated across all corruptions and severity levels with a mIOU of 42.1 compared to both IBE 40.3 and the Softmax Baseline 37.5.

关键词： Robustness Semantics Entropy Training Estimation Task analysis Machine learning

来源：评论

学校读者我要写书评

暂无评论

S6: Semi-Supervised Self-Supervised Semantic Segmentation

S6: Semi-Supervised Self-Supervised Semantic Segmentation

引用

IEEE International Conference on image processing

作者： Moamen Soliman Charles Lehman Ghassan AlRegib OLIVES at the Center for Signal and Information Processing (CSIP) School of Electrical and Computer Engineering Georgia Institute of Technology GA Atlanta

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

Semi-supervised learning provides a means to leverage unlabeled data when labels are expensive to obtain. In this work, we propose a constrained framework that better learns from unlabeled data. The proposed algorithm adds a self-supervised task, image reconstruction, to the target segmentation task. The extra reconstruction task improves the model's geometric reasoning about different textures in an image. It happens that this improvement is transferable from reconstruction to segmentation since they both share some common parts of the architecture. Such extra task allows the trained model to have richer representations and better geometric understanding. Our results show that the proposed constrained framework achieves an improvement in mean Intersection over Union by 18% over unconstrained one, using only 2% of labeled examples. The performance gain and reduction in amounts of labeled data is crucial for applications in which obtaining labels is expensive and labor intensive such as Biomedical Imaging and Seismic Interpretation.

关键词： image reconstruction image segmentation Task analysis Data models Decoding Training Semisupervised learning

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnosis of Wind Energy Conversion Systems Using Gaussian Process Regression-based Multi-Class Random Forest

引用

IFAC-PapersOnLine 2022年第6期55卷 127-132页

作者： Majdi Mansouri Radhia Fezai Mohamed Trabelsi Hajji Mansour Hazem Nounou Mohamed Nounou Electrical and Computer Engineering Program Texas A&M University at Qatar Doha Qatar Department of Mathematics and Sciences Prince Sultan University Riyadh 11586 Saudi Arabia Research Laboratory of Automation Signal Processing and Image National Engineering School of Monastir 5019 Tunisia Electronic and Communications Engineering Department Kuwait College of Science and Technology P. O. Box 27235 Kuwait Higher Institute of Applied Science and Technology of Kasserine University of Kairouan PO Box 471 1200 Kasserine Tunisia Chemical Engineering Program Texas A&M University at Qatar Doha Qatar

This work proposes a new fault diagnosis approach for a wind energy conversion (WEC) system. The proposed technique merges the benefits of feature extraction based on Gaussian Process Regression (GPR) and Multi-Class Random Forest (MCRF)-based fault classification where instances are classified into one or more classes. In the developed GPR-MCRF approach, the nonlinear statistical features including the mean vector M GPR and the variance matrix C GPR are computed using the GPR model with aim of extracting the most relevant features from the WEC system. Then, these features are introduced to the RF classifier for classification and diagnosis purposes. Therefore, the application of the GPR-MCRF technique for WEC systems aims to enhance the use of the classical raw data-based MCRF and diagnosis accuracy. Three kinds of faults (wear-out, open-circuit, and short-circuit faults) are considered in this work. Different case studies are investigated in order to illustrate the effectiveness and robustness of the developed technique compared to the state-of-the-art methods. The obtained results show that the the developed GPR-MCRF technique is an effective feature extraction and fault diagnosis technique for WEC systems.

关键词： Random Forest (RF) Multi-Class Gaussian Process Regression (GPR) Fault Diagnosis Wind Energy Conversion Systems

来源：评论

学校读者我要写书评

暂无评论

Robustness And Overfitting Behavior Of Implicit Background Models

Robustness And Overfitting Behavior Of Implicit Background M...

引用

IEEE International Conference on image processing

作者： Shirley Liu Charles Lehman Ghassan AlRegib OLIVES at the Center for Signal and Information Processing (CSIP) School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA USA

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

In this paper, we examine the overfitting behavior of image classification models modified with Implicit Background Estimation (SCrIBE), which transforms them into weakly supervised segmentation models that provide spatial domain visualizations without affecting performance. Using the segmentation masks, we derive an overfit detection criterion that does not require testing labels. In addition, we assess the change in model performance, calibration, and segmentation masks after applying data augmentations as overfitting reduction measures and testing on various types of distorted images.

关键词： Testing Training Data models image segmentation Transforms Calibration Numerical models

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：