检索结果-内蒙古大学图书馆

Endo-FASt3r: Endoscopic Foundation model Adaptation for Structure from motion

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zeinoddin, Mona Sheikh Islam, Mobarakol Tandogdu, Zafer Shaw, Greg Clarkson, Mathew J. Mazomenos, Evangelos Stoyanov, Danail UCL Hawkes Institute University College London United Kingdom Institute of Health Informatics University College London United Kingdom Dept of Urology University College London Hospitals United Kingdom Dept of Medical Physics & Biomedical Engineering University College London United Kingdom Dept of Computer Science University College London United Kingdom

Accurate depth and camera pose estimation is essential for achieving high-quality 3D visualisations in robotic-assisted surgery. Despite recent advancements in foundation model adaptation to monocular depth estimation of endoscopic scenes via self-supervised learning (SSL), no prior work has explored their use for pose estimation. These methods rely on low rank-based adaptation approaches, which constrain model updates to a low-rank space. We propose Endo-FASt3r, the first monocular SSL depth and pose estimation framework that uses foundation models for both tasks. We extend the Reloc3r relative pose estimation foundation model by designing Reloc3rX, introducing modifications necessary for convergence in SSL. We also present DoMoRA, a novel adaptation technique that enables higher-rank updates and faster convergence. Experiments on the SCARED dataset show that Endo-FASt3r achieves a substantial 10% improvement in pose estimation and a 2% improvement in depth estimation over prior work. Similar performance gains on the Hamlyn and StereoMIS datasets reinforce the generalisability of EndoFASt3r across different datasets. Our code is available at: Endo-FASt3r. Copyright © 2025, The Authors. All rights reserved.

关键词： Endoscopy

Goal-Aware RSS for Complex Scenarios via Program Logic

学校读者我要写书评

暂无评论

Goal-Aware RSS for Complex Scenarios via Program Logic

IEEE Symposium on Intelligent Vehicle

作者： Ichiro Hasuo Clovis Eberhart James Haydon Jérémy Dubut Rose Bohrer Tsutomu Kobayashi Sasinee Pruekprasert Xiao-Yi Zhang Erik André Pallas Akihisa Yamada Kohei Suenaga Fuyuki Ishikawa Kenji Kamijo Yoshiyuki Shinya Takamasa Suetomi National Institute of Informatics (NII) Tokyo Japan SOKENDAI (The Graduate University for Advanced Studies) Japan Japanese-French Laboratory for Informatics (IRL 3527) Tokyo Japan Dept. Computer Science Worcester Polytechnic Institute Worcester MA USA Inst. Software & Systems Engineering University of Augsburg Augsburg Germany Cyber Physical Security Research Center AIST Tokyo Japan Graduate School of Informatics Kyoto University Kyoto Japan Mazda Motor Corporation Fuchu Japan

ISBN: (数字)9798350348811

ISBN: (纸本)9798350348828

We introduce a goal-aware extension of responsibility-sensitive safety (RSS), a recent methodology for rule-based safety guarantee for automated driving systems (ADS). Making RSS rules guarantee goal achievement—in addition to collision avoidance as in the original RSS—requires complex planning over long sequences of manoeuvres. To deal with the complexity, we introduce a compositional reasoning framework based on program logic, in which one can systematically develop RSS rules for smaller subscenarios and combine them to obtain RSS rules for bigger scenarios. As the basis of the framework, we introduce a program logic dFHL that accommodates continuous dynamics and safety conditions. Our framework presents a dFHL-based workflow for deriving goal-aware RSS rules; we discuss its software support, too. We conducted experimental evaluation using RSS rules in a safety architecture. Its results show that goal-aware RSS is indeed effective in realising both collision avoidance and goal achievement.

关键词： Intelligent vehicles computer architecture Software Cognition Safety Planning Complexity theory

Channel Attention HalfU-Net for Skin Lesion Segmentation

学校读者我要写书评

暂无评论

Channel Attention HalfU-Net for Skin Lesion Segmentation

Advanced engineering and Technologies (ICONNIC), International Conference on

作者： Heri Prasetyo Abid Ammar Mahdy Abrar Dwi Fairuz Nadhif Jing-Ming Guo Chih-Hsien Hsia Department of Informatics Universitas Sebelas Maret (UNS) Surakarta Indonesia Department of Electrical Engineering National Taiwan Univ. Sci. Technolgy Taipei Taiwan Dept. Computer Sci. Infor. Engineering National Ilan University Yilan Taiwan

U-shaped architecture have gain popularity in the image segmentation task, especially for biomedical image. Regardless of its good performance, U-shaped architecture generally has huge number of parameters. Considering the resource constraint in actual medical environment, the needs lightweight model is essential. In order to overcome this problem, we propose Channel Attention HalfU-Net. The proposed method implement the existing medical image segmentation model named HalfU-Net. HalfU-Net can reduce the computational cost by simplify the decoder part of the UNet. Channel attention is assigned between ghost model for every stage in encoder part to capture more relevant feature. This leads to performance improvement in skin lesion segmentation problem. Despite that our model achieve state-ofthe-arts performance by surpassing U-Net in aspects of DSC and IoU metrics by 0.86% and 3.48% respectively, the number of parameter of our proposed method is more than 200 times less then U-Net.

关键词：

NCIS: Deep Color Gradient Maps Regression and Three-Class Pixel Classification for Enhanced Neuronal Cell Instance Segmentation in Nissl-Stained Histological Images

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Vadori, Valentina Peruffo, Antonella Graïc, Jean-Marie Finos, Livio Corain, Livio Grisan, Enrico Dept. of Computer Science & Informatics London South Bank University United Kingdom Dept. of Comparative Biomedicine & Food Science University of Padova Italy Dept. of Developmental Psychology and Socialisation University of Padova Italy Dept. of Management and Engineering University of Padova Italy

Deep learning has proven to be more effective than other methods in medical image analysis, including the seemingly simple but challenging task of segmenting individual cells, an essential step for many biological studies. Comparative neuroanatomy studies are an example where the instance segmentation of neuronal cells is crucial for cytoarchitecture characterization. This paper presents an end-to-end framework to automatically segment single neuronal cells in Nissl-stained histological images of the brain, thus aiming to enable solid morphological and structural analyses for the investigation of changes in the brain cytoarchitecture. A U-Net-like architecture with an EfficientNet as the encoder and two decoding branches is exploited to regress four color gradient maps and classify pixels into contours between touching cells, cell bodies, or background. The decoding branches are connected through attention gates to share relevant features, and their outputs are combined to return the instance segmentation of the cells. The method was tested on images of the cerebral cortex and cerebellum, outperforming other recent deep-learning-based approaches for the instance segmentation of cells. © 2023, CC BY-NC-ND.

关键词： Cells

Light up that Droid! On the Effectiveness of Static Analysis Features against App Obfuscation for Android Malware Detection

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Molina-Coronado, Borja Ruggia, Antonio Mori, Usue Merlo, Alessio Mendiburu, Alexander Miguel-Alonso, Jose The Dept. of Computer Architecture and Technology University of the Basque Country UPV/EHU Donostia Spain The Dept. of Computer Science and Artificial Intelligence University of the Basque Country UPV/EHU Donostia Spain The Dept. of Informatics Bioengineering Robotics and Sytems Engineering University of Genoa Genoa Italy The CASD - Centre for Advanced Defense Studies Rome Italy

Malware authors have seen obfuscation as the mean to bypass malware detectors based on static analysis features. For Android, several studies have confirmed that many anti-malware products are easily evaded with simple program transformations. As opposed to these works, ML detection proposals for Android leveraging static analysis features have also been proposed as obfuscation-resilient. Therefore, it needs to be determined to what extent the use of a specific obfuscation strategy or tool poses a risk for the validity of ML malware detectors for Android based on static analysis features. To shed some light in this regard, in this article we assess the impact of specific obfuscation techniques on common features extracted using static analysis and determine whether the changes are significant enough to undermine the effectiveness of ML malware detectors that rely on these features. The experimental results suggest that obfuscation techniques affect all static analysis features to varying degrees across different tools. However, certain features retain their validity for ML malware detection even in the presence of obfuscation. Based on these findings, we propose a ML malware detector for Android that is robust against obfuscation and outperforms current state-of-the-art detectors. © 2023, CC BY-NC-SA.

关键词： Static analysis

An optimization-based algorithm for simultaneous shaping of poles and zeros for non-collocated vibration suppression

学校读者我要写书评

暂无评论

IFAC-PapersOnLine 2022年第16期55卷 394-399页

作者： Adrian Saldanha Haik Silm Wim Michiels Tomáš Vyhlídal Department of Computer Science KU Leuven Celestijnenlaan 200A 3001 Leuven Belgium Dept. Instrumentation and Control Engineering Faculty of Mechanical Engineering and Czech Institute of Informatics Robotics and Cybernetics Czech Technical University in Prague Prague 6 Czechia

This article presents a control design method for simultaneous shaping of the poles and zeros of linear time-invariant systems, motivated by the application of non-collocated vibration suppression to flexible multi-body systems. An entire suppression of vibrations at a target mass for a given excitation frequencies can be recast into the problem of assigning zeros of the transfer function from the excitation force to the target mass' position. The design requirement of achieving sufficient damping in the closed loop system combined with suppressing vibrations at the target, leads to the minimization of the spectral abscissa function of the closed loop system as a function of the controller parameters, subject to zero location constraints. These constraints exhibit polynomial dependence on the controller parameters. We present two approaches to solve the optimization problem which are both based on constraint elimination followed by application of an algorithm for non-smooth unconstrained optimization. The design approach is applicable to delay-free models as well as time-delay models of retarded and neutral type. Simulations results illustrate its applicability to a spring-mass-damper system.

关键词： vibration suppression zero placement time-delay systems spectral optimization

A Hospital Healthcare Monitoring System Using Internet of Things Technologies 6

学校读者我要写书评

暂无评论

A Hospital Healthcare Monitoring System Using Internet of Th...

6th South-East Europe Design Automation, computer engineering, computer Networks and Social Media Conference, SEEDA-CECNSM 2021

作者： Karvounis, Evaggelos Vavva, Maria Giannakeas, Nikolaos Tzallas, Alexandros T. Smanis, Ioannis Tsipouras, Markos G. Department of Informatics and Telecommunication University of Ioannina Arta Greece Modihive Research and Development P.C. Thessaloniki Greece Dept. of Electrical and Computer Engineering Univ. of Western Macedonia Kozani Greece

ISBN: (纸本)9781665427425

In any hospital health care monitoring system, it is necessary to constantly monitor the patient's physiological parameters. Previous studies report that important parameters of any patient that have to be monitored in hospital are heart rate, respiratory rate, oxygen saturation, temperature, change in systolic blood pressure, motion, posture and its location. This work presents a monitoring system that has the capability to monitor in real-time the physiological parameters of the patient using a comfortable wearable device. A Wireless Sensor Network in collaboration with multiple wireless relay nodes, are responsible for collecting and sending the signals from the wireless sensors to the base station. The data is stored and processed using intelligent techniques in a cloud-based environment. Early warning alerts are automatically sent to the medical staff allowing them to intervene on time and earlier than when using manual vital sign observations. By grouping patients according to risk, medical experts can continuously monitor patients health status and quickly identify those that need their attention most. © 2021 IEEE

关键词： Hospitals

Optical Evanescent Field for Direct Observation of Sub-100nm Particle Motion on Glass Surface

学校读者我要写书评

暂无评论

Optical Evanescent Field for Direct Observation of Sub-100nm...

Robotics, engineering, Science, and Technology (RESTCON), International Conference on

作者： Panart Khajornrungruang Thitipat Permpatdechakul Keisuke Suzuki Aran Blattler Yutaka Terayama Daiki Goto Pipat Phaisalpanumas Natthaphon Bun-Athuek Amornphun Phunopas Dept. of Intelligent and Control System Faculty of Computer Science and Systems Engineering Mechanical Science and Technology Division Kyushu Institute of Technology Iizuka-Shi Fukuoka Japan Dept. of Interdisciplinary Informatics Graduate School of Computer Science and Systems Engineering Mechanical Information Science and Technology Division Kyushu Institute of Technology Iizuka-Shi Fukuoka Japan Dept. of Production and Robotics Engineering Faculty of Engineering King Mongkut’s University of Technology North Bangkok Bangsue Bangkok Thailand Dept. of Intelligent and Control System Graduate School of Computer Science and Systems Engineering Mechanical Science and Technology Course Kyushu Institute of Technology Iizuka-Shi Fukuoka Japan Dept. of Industiral Engineering Faculty of Engineering King Mongkut’s University of Technology North Bangkok Bangsue Bangkok Thailand Dept. of Mechanical Engineering Technology College of Industrial Technology King Mongkut’s University of Technology North Bangkok Bangsue Bangkok Thailand

To directly investigate the dynamic nanoscale phenomenon on the surface being processed in wet conditions such as precision polishing, and cleaning in semiconductor industrial, an optical method for visualization and observation of each sub-100 nm sized particle that is moving on an interface such as a silica glass surface by applying an evanescent field have been proposing. Subsequently, we developed an experimental apparatus equipped with an optical microscopy system for verifying the moving particle observation method in a laboratory scale. This article introduces some experimentally direct observation results of duplicated wet processes.

关键词：

Controllable speech synthesis by learning discrete phoneme-level prosodic representations

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Ellinas, Nikolaos Christidou, Myrsini Vioni, Alexandra Sung, June Sig Chalamandaris, Aimilios Tsiakoulis, Pirros Mastorocostas, Paris Innoetics Samsung Electronics Greece Mobile Communications Business Samsung Electronics Korea Republic of Dept. of Informatics & Computer Engineering University of West Attica Greece

In this paper, we present a novel method for phoneme-level prosody control of F0 and duration using intuitive discrete labels. We propose an unsupervised prosodic clustering process which is used to discretize phoneme-level F0 and duration features from a multispeaker speech dataset. These features are fed as an input sequence of prosodic labels to a prosody encoder module which augments an autoregressive attention-based text-to-speech model. We utilize various methods in order to improve prosodic control range and coverage, such as augmentation, F0 normalization, balanced clustering for duration and speaker-independent clustering. The final model enables fine-grained phoneme-level prosody control for all speakers contained in the training set, while maintaining the speaker identity. Instead of relying on reference utterances for inference, we introduce a prior prosody encoder which learns the style of each speaker and enables speech synthesis without the requirement of reference audio. We also fine-tune the multispeaker model to unseen speakers with limited amounts of data, as a realistic application scenario and show that the prosody control capabilities are maintained, verifying that the speaker-independent prosodic clustering is effective. Experimental results show that the model has high output speech quality and that the proposed method allows efficient prosody control within each speaker’s range despite the variability that a multispeaker setting introduces. © 2022, CC BY-NC-ND.

关键词： Speech synthesis