检索结果-内蒙古大学图书馆

Exploring biologically inspired mechanisms of adversarial robustness

Neural computing and Applications 2025年第16期37卷 9331-9342页

作者： Holzhausen, Konstantin Merlid, Mia Torvik, Håkon Olav Malthe-Sørenssen, Anders Lepperød, Mikkel Elle Department of Physics University of Oslo Sem Sælands vei 24 Fysikkbygningen Oslo0371 Norway Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Kristian Augusts gate 23 Oslo0164 Norway

Backpropagation-optimized artificial neural networks, while precise, lack robustness, leading to unforeseen behaviors that affect their safety. Biological neural systems do solve some of these issues already. Unlike artificial models, biological neurons adjust connectivity based on neighboring cell activity. Understanding the biological mechanisms of robustness can pave the way toward building trustworthy and safe systems. Robustness in neural representations is hypothesized to correlate with the smoothness of the encoding manifold. Recent work suggests power law covariance spectra, which were observed studying the primary visual cortex of mice, to be indicative of a balanced trade-off between accuracy and robustness in representations. Here, we show that unsupervised local learning models with winner takes all dynamics learn such power law representations, providing upcoming studies a mechanistic model with that characteristic. Our research aims to understand the interplay between geometry, spectral properties, robustness, and expressivity in neural representations. Hence, we study the link between representation smoothness and spectrum by using weight, Jacobian and spectral regularization while assessing performance and adversarial robustness. Our work serves as a foundation for future research into the mechanisms underlying power law spectra and optimally smooth encodings in both biological and artificial systems. The insights gained may elucidate the mechanisms that realize robust neural networks in mammalian brains and inform the development of more stable and reliable artificial systems. © The Author(s) 2025.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Object Localization with Multiplanar Fiducial Markers: Accurate Pose Estimation 11th

Object Localization with Multiplanar Fiducial Markers: Accu...

引用

11th Iberian Conference on Pattern Recognition and Image analysis, IbPRIA 2023

作者： García-Ruiz, Pablo Muñoz-Salinas, Rafael Medina-Carnicer, Rafael Marín-Jiménez, Manuel J. Department of Computing and Numerical Analysis University of Córdoba Córdoba Spain Córdoba Spain

ISBN: (纸本)9783031366154

Accurate object tracking is an important task in applications such as augmented/virtual reality and tracking of surgical instruments. This paper presents a novel approach for object pose estimation using a dodecahedron with pentagonal fiducial markers. Unlike traditional ArUco markers, which are squared, we propose a pentagonal marker that fits better in the dodecahedron figure. Our proposal improves marker detectability and enhances pose estimation with a novel pose refinement algorithm. Our experiments show the system’s performance and the refinement algorithm’s efficacy under different configurations. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Augmented reality

来源：评论

学校读者我要写书评

暂无评论

DeepArUco: Marker Detection and Classification in Challenging Lighting Conditions 11th

DeepArUco: Marker Detection and Classification in Challeng...

引用

11th Iberian Conference on Pattern Recognition and Image analysis, IbPRIA 2023

作者： Berral-Soler, Rafael Muñoz-Salinas, Rafael Medina-Carnicer, Rafael Marín-Jiménez, Manuel J. Department of Computing and Numerical Analysis University of Córdoba Córdoba Spain Córdoba Spain

ISBN: (纸本)9783031366154

Detection of fiducial markers in challenging lighting conditions can be useful in fields such as industry, medicine, or any other setting in which lighting cannot be controlled (e.g., outdoor environments or indoors with poor lighting). However, explicitly dealing with such conditions has not been done before. Hence, we propose DeepArUco, a deep learning-based framework that aims to detect ArUco markers in lighting conditions where the classical ArUco implementation fails. The system is built around Convolutional Neural Networks, performing the job of detecting and decoding ArUco markers. A method to generate synthetic data to train the networks is also proposed. Furthermore, a real-life dataset of ArUco markers in challenging lighting conditions is introduced and used to evaluate our system, which will be made publicly available alongside the implementation. Code available in GitHub: https://***/AVAuco/deeparuco/. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Proxemics-Net: Automatic Proxemics Recognition in Images 11th

Proxemics-Net: Automatic Proxemics Recognition in Images

引用

11th Iberian Conference on Pattern Recognition and Image analysis, IbPRIA 2023

作者： Jiménez-Velasco, Isabel Muñoz-Salinas, Rafael Marín-Jiménez, Manuel J. Department Computing and Numerical Analysis University of Córdoba Córdoba Spain Córdoba Spain

ISBN: (纸本)9783031366154

Proxemics is a branch of anthropology that studies how humans use personal space as a means of nonverbal communication;that is, it studies how people interact. Due to the presence of physical contact between people, in the problem of proxemics recognition in images, we have to deal with occlusions and ambiguities, which complicates the process of recognition. Several papers have proposed different methods and models to solve this problem in recent years. Over the last few years, the rapid advancement of powerful Deep Learning techniques has resulted in novel methods and approaches. So, we propose Proxemics-Net, a new model that allows us to study the performance of two state-of-the-art deep learning architectures, ConvNeXt and Visual Transformers (as backbones) on the problem of classifying different types of proxemics on still images. Experiments on the existing Proxemics dataset show that these deep learning models do help favorably in the problem of proxemics recognition since we considerably outperformed the existing state of the art, with the ConvNeXt architecture being the best-performing backbone. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Correction: Exploring biologically inspired mechanisms of adversarial robustness(Neural computing and Applications, 10.1007/s00521-025-11019-6)

引用

Neural computing and Applications 2025年第16期37卷 9343-9344页

In this article, panel (b) of Fig. 6 was inadvertently printed in black and white instead of colour;the figure should have appeared as shown below. (Figure presented.) Norm of the Jacobian of latent activations h^ (upper panel) and decision landscape (lower panel) of studied models. Similarly to [12], the displayed planes are created by grid sampling from a random plane projection in the model’s input space The original article has been corrected. © The Author(s) 2025.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning Place Cell Representations and Context-Dependent Remapping 38

Learning Place Cell Representations and Context-Dependent Re...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Pettersen, Markus Rogge, Frederik Lepperød, Mikkel Elle Department of Numerical Analysis and Scientific Computing Simula Research Laboratory Kristian Augusts Gate 23 Oslo Norway Department of Biosciences University of Oslo Blinderveien 31 Oslo Norway

Hippocampal place cells are known for their spatially selective firing patterns, which has led to the suggestion that they encode an animal's location. However, place cells also respond to contextual cues, such as smell. Furthermore, they have the ability to remap, wherein the firing fields and rates of cells change in response to changes in the environment. How place cell responses emerge, and how these representations remap is not fully understood. İn this work, we propose a similarity-based objective function that translates proximity in space, to proximity in representation. We show that a neural network trained to minimize the proposed objective learns place-like representations. We also show that the proposed objective is easily extended to include other sources of information, such as context information, in the same way. When trained to encode multiple contexts, networks learn distinct representations, exhibiting remapping behaviors between contexts. The proposed objective is invariant to orthogonal transformations. Such transformations of the original trained representation (e.g. rotationsİ, therefore yield new representations distinct from the original, without explicit relearning, akin to remapping. Our findings shed new light on the formation and encoding properties of place cells, and also demonstrate an interesting case of representational reuse. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Modeling stochastic eye tracking data: A comparison of quantum generative adversarial networks and Markov models

Modeling stochastic eye tracking data: A comparison of quant...

引用

2024 Genetic and Evolutionary Computation Conference Companion, GECCO 2024 Companion

作者： Bhandari, Shailendra Lencastre, Pedro Lind, Pedro Department of Computer Science OsloMet - Oslo Metropolitan University OsloN-0130 Norway Simula Research Laboratory Numerical Analysis and Scientific Computing Oslo0164 Norway

ISBN: (纸本)9798400704956

We explore the use of quantum generative adversarial networks QGANs for modeling eye movement velocity data. We assess whether the advanced computational capabilities of QGANs can enhance the modeling of complex stochastic distribution beyond the traditional mathematical models, particularly the Markov model. The findings indicate that while QGANs demonstrate potential in approximating complex distributions, the Markov model consistently outperforms in accurately replicating the real data distribution. This comparison underlines the challenges and avenues for refinement in time series data generation using quantum computing techniques. It emphasizes the need for further optimization of quantum models to better align with real-world data characteristics. © 2024 is held by the owner/author(s).

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

SCALABLE APPROXIMATION AND SOLVERS FOR IONIC ELECTRODIFFUSION IN CELLULAR GEOMETRIES

arXiv

引用

arXiv 2024年

作者： Benedusi, Pietro Ellingsrud, Ada J. Herlyng, Halvor Rognes, Marie E. Department for Numerical Analysis and Scientific Computing Simula Research Laboratory Oslo Norway

The activity and dynamics of excitable cells are fundamentally regulated and moderated by extracellular and intracellular ion concentrations and their electric potentials. The increasing availability of dense reconstructions of excitable tissue at extreme geometric detail pose a new and clear scientific computing challenge for computational modelling of ion dynamics and transport. In this paper, we design, develop and evaluate a scalable numerical algorithm for solving the time-dependent and nonlinear KNP-EMI equations describing ionic electrodiffusion for excitable cells with an explicit geometric representation of intracellular and extracellular compartments and interior interfaces. We also introduce and specify a set of model scenarios of increasing complexity suitable for benchmarking. Our solution strategy is based on an implicit-explicit discretization and linearization in time, a mixed finite element discretization of ion concentrations and electric potentials in intracellular and extracellular domains, and an algebraic multigrid-based, inexact block-diagonal preconditioner for GMRES. numerical experiments with up to 108 unknowns per time step and up to 256 cores demonstrate that this solution strategy is robust and scalable with respect to the problem size, time discretization and number of *** Codes 65F10, 65N30, 65N55, 68U20, 68W10, 92-08, 92C20 © 2024, CC BY-SA.

关键词： Electric potential

来源：评论

学校读者我要写书评

暂无评论

analysis on the Rejection and Device Passivity Problems with Myoelectric Prosthetic Hand Control 28

Analysis on the Rejection and Device Passivity Problems with...

引用

28th World Multi-Conference on Systemics, Cybernetics and Informatics, WMSCI 2024

作者： Al-Owaidi, Noora Mora, Marta C. Ventura, Sebastián Department of Computing and Numerical Analysis University of Cordoba Córdoba14071 Spain Department of Mechanical Engineering and Construction Universitat Jaume I Castellón12071 Spain University of Cordoba Córdoba14071 Spain

ISBN: (纸本)9781950492794

Despite recent technological advancements in the field of intelligent prosthetic hands, there remains a high rejection rate among patients. The reasons for this rejection are diverse, ranging from limited functionality, and perplexing interfaces, to discomfort, with most issues being discernible only through subjective self-assessment. Surprisingly, there is a lack of specific methodologies to gauge the superiority of a new prosthetic solution over its predecessors. It is imperative to delve into the complexities of maneuvering myoelectric prosthetic hands to comprehend the challenges faced by users and enhance prosthetic accessories and rehabilitation techniques. These hurdles may lead to either a passive usage or a complete abandonment of the prosthetic device. To surmount these obstacles, the field of prosthetics incessantly seeks more sophisticated technologies to enhance functionality, user-friendliness, and device longevity, and reduce maintenance. This study undertakes a methodical examination of the literature on control issues concerning myoelectric prosthetic hands, encompassing recent advancements in sensor technology, hurdles in signal quality, user training and adaptation, issues with pattern recognition algorithms, user satisfaction and comfort, battery life and maintenance, psychological aspects, expenses, and dexterity, culminating in an exploration of functional challenges, advantages, consequences, and feedback. © 2024 by the International Institute of Informatics and Systemics. All rights reserved.

关键词： Sensory feedback

来源：评论

学校读者我要写书评

暂无评论

A SPLITTING, DISCONTINUOUS GALERKIN SOLVER FOR THE CELL-BY-CELL ELECTRONEUTRAL NERNST-PLANCK FRAMEWORK

arXiv

引用

arXiv 2024年

作者： Ellingsrud, Ada J. Kuchta, Miroslav Department for Numerical Analysis and Scientific Computing Simula Research Laboratory Oslo Norway

Mathematical models for excitable tissue with explicit representation of individual cells are highly detailed and can, unlike classical homogenized models, represent complex cellular geometries and local membrane variations. However, these cell-based models are challenging to approximate numerically, partly due to their mixed-dimensional nature with unknowns both in the bulk and at the lower-dimensional cellular membranes. We here develop and evaluate a novel solution strategy for the cell-based KNP-EMI model describing ionic electrodiffusion in and between intra- and extracellular compartments with explicit representation of individual cells. The strategy is based on operator splitting, a multiplier-free formulation of the coupled dynamics across sub-regions, and a discontinuous Galerkin discretization. In addition to desirable theoretical properties, such as local mass conservation, the scheme is practical as it requires no specialized functionality in the finite element assembly and order optimal solvers for the resulting linear systems can be realized with black-box algebraic multigrid preconditioners. numerical investigations show that the proposed solution strategy is accurate, robust with respect to discretization parameters, and that the parallel scalability of the solver is close to optimal – both for idealized and realistic two and three dimensional *** Codes 65M60, 65F10, 65M55, 68U20, 92-08, 92C37 © 2024, CC BY.

关键词： Linear systems

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：