检索结果-内蒙古大学图书馆

2023 IEEE/CVF Conference on computer Vision and Pattern Recognition Workshops, CVPRW 2023

作者： Akash Guna, R.T. Benitez, Raul Sikha, O.K. Amrita School of Engineering Department of Computer Science and Engineering Coimbatore India Universitat Politècnica de Catalunya BarcelonaTech Departament of Automatic Control Spain

ISBN: (纸本)9798350302493

Existing explainability approaches for convolutional neural networks (CNNs) are mainly applied after training (post-hoc) which is generally unreliable. Ante-hoc explainers trained simultaneously with the CNN are more reliable. However, current ante-hoc explanation methods mainly generate inexplicit concept-based explanations tailored to specific tasks. To address these limitations, we propose a task-agnostic ante-hoc framework that can generate interpretation maps to visually explain any CNN. Our framework simultaneously trains the CNN and a weighting network - an explanation generation module. The generated maps are self-explanatory, eliminating the need for manual identification of concepts. We demonstrate that our method can interpret classification, facial landmark detection, and image captioning tasks. We show that our framework is explicit, faithful, and stable through experiments. To the best of our knowledge, this is the first ante-hoc CNN explanation strategy that produces visual explanations generic to CNNs for different tasks. © 2023 IEEE.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Ante-Hoc Generation of Task-Agnostic Interpretation Maps

Ante-Hoc Generation of Task-Agnostic Interpretation Maps

引用

IEEE computer Society Conference on computer Vision and Pattern Recognition Workshops (CVPRW)

作者： Akash Guna R T Raul Benitez Sikha O K Department of Computer Science and Engineering Amrita School of Engineering Coimbatore India Departament of Automatic Control Universitat Politècnica de Catalunya-BarcelonaTech Spain

关键词：

来源：评论

学校读者我要写书评

暂无评论

design of switching state-feedback controllers for linear systems subject to asymmetric saturations *

引用

IFAC-PapersOnLine 2023年第2期56卷 3265-3270页

作者： Adrián Ruiz Damiano Rotondo Bernardo Morcego Research Center for Supervision Safety and Automatic Control Universitat Politècnica de Catalunya (UPC) Rambla Sant Nebridi 22 08222 Terrassa Spain Department of Electrical and Computer Engineering (IDE) University of Stavanger (UiS) Kristine Bonnevies vei 22 4021 Stavanger Norway

This paper considers the problem of controlling a linear system affected by asymmetrical input saturation. The proposed solution is based on using a linear matrix inequality (LMI)-based methodology to find the gains of a switching state-feedback controller. The main difference and contribution when compared to existing approaches is that the switching rule is chosen based on the closed-loop performance that each of the non-saturating controller gains can achieve when used with the current value of the state vector. Although the main focus of the paper is on time-invariant systems, the possible extension to linear parameter-varying (LPV) systems is discussed. An illustrative example is used to show the main features of the proposed approach.

关键词： Linear systems Saturation Linear matrix inequalities (LMIs)

来源：评论

学校读者我要写书评

暂无评论

IARS SegNet: Interpretable Attention Residual Skip connection SegNet for melanoma segmentation

arXiv

引用

arXiv 2023年

作者： Narayanan, Shankara Sikha, O.K. Benitez, Raul Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Coimbatore641112 India Department of Automatic Control Universitat Politècnica de Catalunya Av. d’Eduard Maristany Barcelona08019 Spain Barcelona08034 Spain

Skin lesion segmentation plays a crucial role in the computer-aided diagnosis of melanoma. deep Learning models have shown promise in accurately segmenting skin lesions, but their widespread adoption in real-life clinical settings is hindered by their inherent black-box nature. In domains as critical as healthcare, interpretability is not merely a feature but a fundamental requirement for model adoption. This paper proposes IARS SegNet an advanced segmentation framework built upon the SegNet baseline model. Our approach incorporates three critical components: Skip connections, residual convolutions, and a global attention mechanism onto the baseline Segnet architecture. These elements play a pivotal role in accentuating the significance of clinically relevant regions, particularly the contours of skin lesions. The inclusion of skip connections enhances the model’s capacity to learn intricate contour details, while the use of residual convolutions allows for the construction of a deeper model while preserving essential image features. The global attention mechanism further contributes by extracting refined feature maps from each convolutional and deconvolutional block, thereby elevating the model’s interpretability. This enhancement highlights critical regions, fosters better understanding, and leads to more accurate skin lesion segmentation for melanoma diagnosis. This study primarily focuses on the interpretation of performance improvements in the base model resulting from the integration of each of these three components. To comprehensively assess the performance gain achieved with each addition, we employ two sets of evaluation metrics, quantifying performance based on both regions and contours. The results underscore the superior segmentation capabilities of the proposed architecture compared to the SegNet and U-Net models. Notably, it provides interpretable results, particularly when applied to the PH2 dataset. Copyright © 2023, The Authors. All rights reserved

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Learning from demonstrations for autonomous soft-tissue retraction

arXiv

引用

arXiv 2021年

作者： Pore, Ameya Tagliabue, Eleonora Piccinelli, Marco Dall’Alba, Diego Casals, Alicia Fiorini, Paolo Department of Computer Science University of Verona Italy Automatic Control and Computer Engineering Department Universitat Politècnica de Catalunya Barcelona Spain

The current research focus in Robot-Assisted Minimally Invasive Surgery (RAMIS) is directed towards increasing the level of robot autonomy, to place surgeons in a supervisory position. Although Learning from demonstrations (LfD) approaches are among the preferred ways for an autonomous surgical system to learn expert gestures, they require a high number of demonstrations and show poor generalization to the variable conditions of the surgical environment. In this work, we propose an LfD methodology based on Generative Adversarial Imitation Learning (GAIL) that is built on a deep Reinforcement Learning (DRL) setting. GAIL combines generative adversarial networks to learn the distribution of expert trajectories with a DRL setting to ensure generalisation of trajectories providing human-like behaviour. We consider automation of tissue retraction, a common RAMIS task that involves soft tissues manipulation to expose a region of interest. In our proposed methodology, a small set of expert trajectories can be acquired through the da Vinci Research Kit (dVRK) and used to train the proposed LfD method inside a simulated environment. Results indicate that our methodology can accomplish the tissue retraction task with human-like behaviour while being more sample-efficient than the baseline DRL method. Towards the end, we show that the learnt policies can be successfully transferred to the real robotic platform and dep.oyed for soft tissue retraction on a synthetic phantom. © 2021, CC BY.

关键词： deep learning

来源：评论

学校读者我要写书评

暂无评论

Observer Attack on Stream Ciphers

引用

IFAC-PapersOnLine 2022年第30期55卷 13-18页

作者： Ramachandran Anantharaman Virendra Sule Research Center for Supervision Safety and Automatic Control Universitat Politècnica de Catalunya (UPC) Rambla Sant Nebridi 22 08222 Terrassa Spain Department of Electrical and Computer Engineering (IDE) University of Stavanger (UiS) Kristine Bonnevies vei 22 4021 Stavanger Norway

This paper proposes an application of a new observer theory for non-linear systems developed previously to solve the Cryptanalysis problem of a special class of pseudorandom generators which are commonly used in Cryptography. The Cryptanalysis problem addressed here is that of the recovery of internal state of the non-linear dynamic stream generator from the output stream. The proposed methodology is termed as observability attack. It is also shown that for a special class of generators, the computations are of complexity O (D 4 ) in pre-computation and of O(D) for online computation, where D  = σ d i=0 (n i) for this class of stream generators with n states and d the degree of the output function. The attack is technically applicable over general finite fields as well as most dynamic systems arising from models of stream ciphers and appropriate bounds on computation are estimated. From these complexity bounds, it follows that this attack is feasible in realistic cases and gives important estimates of time and memory resources required for Cryptanalysis of a class of stream ciphers.

关键词： Stream ciphers Cryptanalysis Boolean Dynamical Systems Observer design Koopman Operator

来源：评论

学校读者我要写书评

暂无评论

Semi-supervised active transfer learning for fetal ECG arrhythmia detection

引用

computer Methods and Programs in Biomedicine Update 2023年 3卷

作者： Mohebbian, Mohammad Reza Marateb, Hamid Reza Wahid, Khan A. Department of Electrical and Computer Engineering University of Saskatchewan Saskatoon S7N 5A9 SK Canada Biomedical Engineering Department Engineering Faculty University of Isfahan Isfahan Iran Biomedical Engineering Research Centre (CREB) Automatic Control Department (ESAII) Universitat Politècnica de Catalunya-Barcelona Tech (UPC) Barcelona Spain

deep learning has demonstrated excellent results for ECG anomaly detection, wherein most approaches used supervised learning. The requirement of thousands of manually annotated samples is a concern for state-of-the-art anomaly detection systems, especially for fetal ECG (FECG), and currently, there is not a publicly available FECG dataset annotated for each FECG beat. In this paper, we offer a modified active learning technique based on transfer learning, calibration probability, and autoencoder-based sampling to reduce number of samples requires to annotate. In this regard, we used 25,000 s of recording from 47 patients from the MIT-BIH Arrhythmia Database to train a deep learning model to detect anomalies in non-fetus subjects. Then we used the unlabeled Non-Invasive Fetal ECG Arrhythmia Database (NIFEA DB) of 26 subjects to fine-tune the trained model to fine-tune the trained model based on active learning to detect anomalies in binary form for fetal. A variational autoencoder is trained on all data (adult and fetal ECG), and clustering is applied to latent features extracted from data after dimension reduction. Then, the sampling process of active learning selected samples from different clusters with low confidence to cover all data distribution. Moreover, a probability calibration based on mc-dropout and isotonic regression is used to calibrate confidences, helping to select reliable low-confidence samples. Various ablation studies were performed to show the influence of autoencoder-based sampling, calibration, and transfer learning, which showed that the proposed method could achieve 92% accuracy using 399 training samples. In contrast, other methods required more training samples to reach the same level of accuracy without calibration or an autoencoder and clustering approach or training without active learning. The study also found that transfer learning significantly impacted faster convergence and that the proposed active learning approach was more effect

关键词： Active learning Anomaly Arrhythmia FECG Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Extremum Seeking control for Beam Steering using Hypersurfaces

Extremum Seeking Control for Beam Steering using Hypersurfac...

引用

IEEE International Conference on Communications Workshops, ICC

作者： Nouman Ashraf Marios Lestas Taqwa Saeed Hamidreza Taghvaee Sergi Abadal Andreas Pitsillides Christos Liaskos Dep. of Electrical Engineering Frederick University Dep. of Computer Science University of Cyprus Nicosia Cyprus Dep. of Computer Architecture Universitat Politècnica de Catalunya Barcelona Spain Institute of Computer Science Foundation of Research and Technology Hellas Heraklion Greece

ISBN: (数字)9781728174402

ISBN: (纸本)9781728174419

Hypersurface, a software defined metasurface (SDM) paradigm constitutes a revolutionary technology aiming at offering programmatic control over all aspects of a propagating wave, altering its direction, power, polarization and phase. In the absence of line of sight (LOS) between a transmitter and a receiver, HSF can provide seamless connectivity via programmatic reflection. However, such an application requires fine tuning of the metasurface reconfiguration parameters, which may not be effective when done in an open loop fashion due to model uncertainties. In this work, we consider a feedback-based formulation of the problem, which involves maximization of the received power and propose the use of Extremum Seeking control (ESC) for the controller implementation due to fact that it is model-free and adheres to the maximization problem. Extensive simulations indicate that the proposed scheme is successful in guiding the impinging wave to the receiver within reasonable time even in the presence of time varying delays incurred by message propagation. In addition, a discrete time implementation is considered, investigating its limitations as we increase the sampling time while also characterizing the traffic within the controller network.

关键词： Receivers Beam steering Transmitters Aerospace electronics computer science Software delays

来源：评论

学校读者我要写书评

暂无评论

development of animated facial expressions to express emotions in a robot: RobotIcon

Development of animated facial expressions to express emotio...

引用

IEEE Ecuador Technical Chapters Meeting (ETCM)

作者： Lukáš Danev Marten Hamann Nicolas Fricke Tobias Hollarek dennys Paillacho Facultat d‚ Informática de Barcelona Universitat Politécnica de Catalunya (UPC) Barcelona Spain Department of Automatic Control Universitat Politécnica de Catalunya Barcelona Spain Electrical and Computer Engineering Department Escuela Superior Politécnica del Litoral (ESPOL) Guayaquil Ecuador

ISBN: (纸本)9781538638958

Robotics is a field which does not only become more and more important in industry and production but is becoming part of many peoples every day life and as such, the interaction between robots and humans is an essential field of research. In this work, we present a face - being an important part of a natural communication - and show that animated facial expressions can be used to express the robot's emotions. To do so, we developed a set of animated facial expressions to be shown on the screen of the MASHI robot, a telepresence robot with a wide-screen display mounted on top. A survey was conducted to evaluate the recognition value of the facial expressions as well as to evaluate whether displaying the full face or rather just the eye-region affects this recognition rate. The evaluation of this survey shows that most of our emotions are recognized better if a mouth is provided. Yet, the recognition rate is still acceptable without displaying the mouth. In summary, all emotions were well recognized and therefore, displaying these on robots like MASHI will contribute to a more natural human-robot interaction.

关键词： Face Mouth Service robots Emotion recognition Shape Face recognition

来源：评论

学校读者我要写书评

暂无评论

26th Annual Computational Neuroscience Meeting (CNS*2017): Part 3 Antwerp, Belgium. 15-20 July 2017 Abstracts

引用

BMC NEUROSCIENCE 2017年第SUPPL 1期18卷 95-176页

作者： [Anonymous] Department of Neuroscience Yale University New Haven CT 06520 USA Department Physiology & Pharmacology SUNY Downstate Brooklyn NY 11203 USA NYU School of Engineering 6 MetroTech Center Brooklyn NY 11201 USA Departament de Matemàtica Aplicada Universitat Politècnica de Catalunya Barcelona 08028 Spain Institut de Neurobiologie de la Méditerrannée (INMED) INSERM UMR901 Aix-Marseille Univ Marseille France Center of Neural Science New York University New York NY USA Aix-Marseille Univ INSERM INS Inst Neurosci Syst Marseille France Laboratoire de Physique Théorique et Modélisation CNRS UMR 8089 Université de Cergy-Pontoise 95300 Cergy-Pontoise Cedex France Department of Mathematics and Computer Science ENSAT Abdelmalek Essaadi’s University Tangier Morocco Laboratory of Natural Computation Department of Information and Electrical Engineering and Applied Mathematics University of Salerno 84084 Fisciano SA Italy Department of Medicine University of Salerno 84083 Lancusi SA Italy Dipartimento di Fisica Università degli Studi Aldo Moro Bari and INFN Sezione Di Bari Italy Data Analysis Department Ghent University Ghent Belgium Coma Science Group University of Liège Liège Belgium Cruces Hospital and Ikerbasque Research Center Bilbao Spain BIOtech Department of Industrial Engineering University of Trento and IRCS-PAT FBK 38010 Trento Italy Department of Data Analysis Ghent University Ghent 9000 Belgium The Wellcome Trust Centre for Neuroimaging University College London London WC1N 3BG UK Department of Electronic Engineering NED University of Engineering and Technology Karachi Pakistan Blue Brain Project École Polytechnique Fédérale de Lausanne Lausanne Switzerland Departement of Mathematics Swansea University Swansea Wales UK Laboratory for Topology and Neuroscience at the Brain Mind Institute École polytechnique fédérale de Lausanne Lausanne Switzerland Institute of Mathematics University of Aberdeen Aberdeen Scotland UK Department of Integrativ

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：