检索结果-内蒙古大学图书馆

2024 IEEE International Conference on Robotics and Biomimetics, ROBIO 2024

作者： Slim, Malak Daher, Naseem Elhajj, Imad H. American University of Beirut Vision and Robotics Lab Electrical and Computer Engineering Department Beirut Lebanon

ISBN: (纸本)9781665481090

In this work, we present a novel solution aimed at improving robotic manipulators' performance in contact tasks. Inspired by the human motor control system, which relies on a feedforward mechanism to anticipate and plan movements based on the physical properties of the target environment, our approach plans the robot's motion during the reaching phase, prior to contact. To validate our approach, we conducted experiments using the KUKA youBot arm in two distinct environments, represented by soft and hard materials. Results showed that the robot exhibited compliant behavior, with an average reduction of 71% in overshoot, 60% in rise-time, and 68% steady-state error of the force control response during contact. © 2024 IEEE.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

On Learning Deep O(n)-Equivariant Hyperspheres 41

On Learning Deep O(n)-Equivariant Hyperspheres

引用

41st International Conference on Machine Learning, ICML 2024

作者： Melnyk, Pavlo Felsberg, Michael Wadenbäck, Mårten Robinson, Andreas Le, Cuong Computer Vision Laboratory Department of Electrical Engineering Linköping University Sweden

In this paper, we utilize hyperspheres and regular n-simplexes and propose an approach to learning deep features equivariant under the transformations of nD reflections and rotations, encompassed by the powerful group of O(n). Namely, we propose O(n)-equivariant neurons with spherical decision surfaces that generalize to any dimension n, which we call Deep Equivariant Hyperspheres. We demonstrate how to combine them in a network that directly operates on the basis of the input points and propose an invariant operator based on the relation between two points and a sphere, which as we show, turns out to be a Gram matrix. Using synthetic and real-world data in nD, we experimentally verify our theoretical contributions and find that our approach is superior to the competing methods for O(n)-equivariant benchmark datasets (classification and regression), demonstrating a favorable speed/performance trade-off. The code is available on GitHub. Copyright 2024 by the author(s)

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

A framework for locating multiple RFID tags using RF hologram tensors

引用

Digital Communications and Networks 2025年第2期11卷 337-348页

作者： Xiangyu Wang Jian Zhang Shiwen Mao Senthilkumar CG Periaswamy Justin Patton Dept.of Electrical and Computer Engineering Auburn UniversityAuburnAL 36849-5201USA Department of Electrical and Computer Engineering Kennesaw State UniversityKennesawGA 30144USA RFID Lab Auburn UniversityAuburnAL 36849USA

In this paper,we present a Deep Neural Network(DNN)based framework that employs Radio Frequency(RF)hologram tensors to locate multiple Ultra-High Frequency(UHF)passive Radio-Frequency Identification(RFID)*** RF hologram tensor exhibits a strong relationship between observation and spatial location,helping to improve the robustness to dynamic environments and *** RFID data is often marred by noise,we implement two types of deep neural network architectures to clean up the RF hologram *** the spatial relationship between tags,the deep networks effectively mitigate fake peaks in the hologram tensors resulting from multipath propagation and phase *** contrast to fingerprinting-based localization systems that use deep networks as classifiers,our deep networks in the proposed framework treat the localization task as a regression problem preserving the ambiguity between *** also present an intuitive peak finding algorithm to obtain estimated locations using the sanitized hologram *** proposed framework is implemented using commodity RFID devices,and its superior performance is validated through extensive experiments.

关键词： Radio-frequency identification(RFID) Ultra-high frequency(UHF)passive RFID tag RF hologram tensor Indoor localization Deep learning(DL) Swin Transformer Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Graph-Segmenter:graph transformer with boundary-aware attention for semantic segmentation

引用

Frontiers of computer Science 2024年第5期18卷 97-108页

作者： Zizhang WU Yuanzhu GAN Tianhao XU Fan WANG Computer Vision Perception Department of ZongMu Technology Shanghai 201203China Faculty of Electrical Engineering Information TechnologyPhysicsTechnical University of BraunschweigBraunschweig 38106Germany

Thetransformer-based semantic segmentation approaches,which divide the image into different regions by sliding windows and model the relation inside each window,have achieved outstanding ***,since the relation modeling between windows was not the primary emphasis of previous work,it was not fully *** address this issue,we propose a Graph-Segmenter,including a graph transformer and a boundary-aware attention module,which is an effective network for simultaneously modeling the more profound relation between windows in a global view and various pixels inside each window as a local one,and for substantial low-cost boundary ***,we treat every window and pixel inside the window as nodes to construct graphs for both views and devise the graph *** introduced boundary-awareattentionmoduleoptimizes theedge information of the target objects by modeling the relationship between the pixel on the object's *** experiments on three widely used semantic segmentation datasets(Cityscapes,ADE-20k and PASCAL Context)demonstrate that our proposed network,a Graph Transformer with Boundary-aware Attention,can achieve state-of-the-art segmentation performance.

关键词： graph transformer graph relation network boundary-aware attention semantic segmentation

来源：评论

学校读者我要写书评

暂无评论

A transformer model for boundary detection in continuous sign language

引用

Multimedia Tools and Applications 2024年第42期83卷 89931-89948页

作者： Rastgoo, Razieh Kiani, Kourosh Escalera, Sergio Electrical and Computer Engineering Faculty Semnan University Semnan*** Iran Sergio Escalera Department of Mathematics and Informatics Universität de Barcelona and Computer Vision Center Barcelona Spain

Sign Language Recognition (SLR) has garnered significant attention from researchers in recent years, particularly the intricate domain of Continuous Sign Language Recognition (CSLR), which presents heightened complexity compared to Isolated Sign Language Recognition (ISLR). One of the prominent challenges in CSLR pertains to accurately detecting the boundaries of isolated signs within a continuous video stream. Additionally, the reliance on handcrafted features in existing models poses a challenge to achieving optimal accuracy. To surmount these challenges, we propose a novel approach utilizing a Transformer-based model. Unlike traditional models, our approach focuses on enhancing accuracy while eliminating the need for handcrafted features. The Transformer model is employed for both ISLR and CSLR. The training process involves using isolated sign videos, where hand keypoint features extracted from the input video are enriched using the Transformer model. Subsequently, these enriched features are forwarded to the final classification layer. The trained model, coupled with a post-processing method, is then applied to detect isolated sign boundaries within continuous sign videos. The evaluation of our model is conducted on two distinct datasets, including both continuous signs and their corresponding isolated signs, demonstrates promising results. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

A New Multi-Source Light Detection Benchmark and Semi-Supervised Focal Light Detection 38

A New Multi-Source Light Detection Benchmark and Semi-Superv...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Baek, Jae-Yong Yoo, Yong-Sang Bae, Seung-Hwan Inha University Dept. of Electrical and Computer Engineering Vision and Learning Lab Korea Republic of

This paper addresses a multi-source light detection (LD) problem from vehicles, traffic signals, and streetlights under driving scenarios. Albeit it is crucial for autonomous driving and night vision, this problem has not been yet focused on as much as other object detection (OD). One of the main reasons is the absence of a public available LD benchmark dataset. Therefore, we construct a new large LD dataset consisting of different light sources via heavy annotation:YouTube Driving Light Detection dataset (YDLD). Compared to the existing LD datasets, our dataset has much more images and box annotations for multi-source lights. We also provide rigorous statistical analysis and transfer learning comparison of other well-known detection benchmark datasets to prove the generality of our YDLD. For the recent object detectors, we achieve the extensive comparison results on YDLD. However, they tend to yield the low mAP scores due to the intrinsic challenges of LD caused by very tiny size and similar appearance. To resolve those, we design a novel lightness focal loss which penalizes miss-classified samples more and a lightness spatial attention prior by reflecting a global scene context. In addition, we develop a semi-supervised focal light detection (SS-FLD) by embedding our lightness focal loss into the semi-supervised object detection (SSOD). We prove that our methods can consistently boost mAP to the variety of types of recent detectors on YDLD. We will open both YDLD and SS-FLD code at https://***/YDLD-dataset/YDLD. © 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

Local Binary Pattern (LBP) and Transfer Learning Based Approach to Classify Lung and Colon Cancer

引用

SN computer Science 2024年第6期5卷 783页

作者： Singh, Onkar Singh, Koushlendra Kumar Department of Computer Science and Engineering Manipal University Jaipur Machine Vision and Intelligence Lab Department of Computer Science and Engineering National Institute of Technology Jamshedpur

Several genetic disorders and other metabolic abnormalities work together to generate the lethal disease known as cancer. Today’s most contributing factors to mortality and disability in patients are lung and colon cancer. A World Health Organization (WHO) 2020 report listed cancer as one of the leading causes of mortality globally. About 2.735 million of these fatalities were caused by lung and colon cancer combined. A critical component of the patient’s treatment is the diagnosis of lung cancer by histopathology. Therefore, one of the leading research priorities, mostly in the domain of biomedical health information systems, is the identification and categorization of lung and colon cancer. The present article encompasses the Local Binary Pattern (LBP) and transfer learning-based approaches to classify lung as well as colon cancer. LBP has been used for extracting features, and transfer learning has been used for the classification of lung and colon cancer. Histopathological (LC25000) lung and colon datasets are used to validate the proposed methodology. The results of the proposed method have also been compared with different existing methods and reported in the article. Our proposed method has an average accuracy of 99.00% and a F1 score of 99.2%, whereas precision and recall have 99.4% for lung and colon cancer detection. The results of the investigation demonstrate that our suggested technique greatly outperforms current models. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2024.

关键词： Histopathology images Local binary pattern (LBP) Lung cancer Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Quantitative phase imaging endoscopy with a metalens

引用

Light(Science & Applications) 2024年第12期13卷 3196-3209页

作者： Aamod Shanker Johannes E.Fröch Saswata Mukherjee Maksym Zhelyeznyakov Steven L.Brunton Eric J.Seibel Arka Majumdar Department of Electrical and Computer Engineering University of WashingtonSeattleWA 98195USA Center for Vision Science University of RochesterNew YorkNY 14623USA Department of Mechanical Engineering University of WashingtonSeattleWA 98105USA Department of Physics University of WashingtonSeattle 98195USA

Quantitative phase imaging(QPI)recovers the exact wavefront of light from intensity *** and optical density maps of translucent microscopic bodies can be extracted from these quantified phase *** demonstrate quantitative phase imaging at the tip of a coherent fiber bundle using chromatic aberrations inherent in a silicon nitride hyperboloid *** method leverages spectral multiplexing to recover phase from multiple defocus planes in a single capture using a color *** 0.5mm aperture metalens shows robust quantitative phase imaging capability with a 28°field of view and 0.2πphase resolution(~0.1λin air)for experiments with an endoscopic fiber *** the spectral functionality is encoded directly in the imaging lens,the metalens acts both as a focusing element and a spectral *** use of a simple computational backend will enable real-time *** limitations in the adoption of phase imaging methods for endoscopy such as multiple acquisition,interferometric alignment or mechanical scanning are completely mitigated in the reported metalens based QPI.

关键词： phase bundle exact

来源：评论

学校读者我要写书评

暂无评论

Class Activation Mapping and Uncertainty Estimation in Multi-Organ Segmentation

Class Activation Mapping and Uncertainty Estimation in Multi...

引用

Medical Imaging 2023: computer-Aided Diagnosis

作者： Sadique, M.S. Farzana, W. Temtam, A. Iftekharuddin, K.M. Vision Lab Department of Electrical and Computer Engineering Old Dominion University NorfolkVA23529 United States

ISBN: (纸本)9781510660359

Deep learning (DL)-based medical imaging and image segmentation algorithms achieve impressive performance on many benchmarks. Yet the efficacy of deep learning methods for future clinical applications may become questionable due to the lack of ability to reason with uncertainty and interpret probable areas of failures in prediction decisions. Therefore, it is desired that such a deep learning model for segmentation classification is able to reliably predict its confidence measure and map back to the original imaging cases to interpret the prediction decisions. In this work, uncertainty estimation for multiorgan segmentation task is evaluated to interpret the predictive modeling in DL solutions. We use the state-of-the-art nnU-Net to perform segmentation of 15 abdominal organs (spleen, right kidney, left kidney, gallbladder, esophagus, liver, stomach, aorta, inferior vena cava, pancreas, right adrenal gland, left adrenal gland, duodenum, bladder, prostate/uterus) using 200 patient cases for the Multimodality Abdominal Multi-Organ Segmentation Challenge 2022. Further, the softmax probabilities from different variants of nnU-Net are used to compute the knowledge uncertainty in the deep learning framework. Knowledge uncertainty from ensemble of DL models is utilized to quantify and visualize class activation map for two example segmented organs. The preliminary result of our model shows that class activation maps may be used to interpret the prediction decision made by the DL model used in this study. © 2023 SPIE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Adaptive Critic Network for Person Tracking using 3D Skeleton Data 34

Adaptive Critic Network for Person Tracking using 3D Skeleto...

引用

Pattern Recognition and Tracking XXXIV 2023

作者： Zalameda, Joseph G. Glandon, Alex Iftekharuddin, Khan M. Vision Lab in Department of Electrical and Computer Engineering Old Dominion University NorfolkVA23529 United States

ISBN: (数字)9781510661691

ISBN: (纸本)9781510661684

Analysis of human gait using 3-dimensional co-occurrence skeleton joints extracted from Lidar sensor data has been shown a viable method for predicting person identity. The co-occurrence based networks rely on the spatial changes between frames of each joint in the skeleton data sequence. Normally, this data is obtained using a Lidar skeleton extraction method to estimate these co-occurrence features from raw Lidar frames, which can be prone to incorrect joint estimations when part of the body is occluded. These datasets can also be time consuming and expensive to collect and typically offer a small number of samples for training and testing network models. The small number of samples and occlusion can cause challenges when training deep neural networks to perform real time tracking of the person in the scene. We propose preliminary results with a deep reinforcement learning actor critic network for person tracking of 3D skeleton data using a small dataset. The proposed approach can achieve an average tracking rate of 68.92±15.90% given limited examples to train the network. © 2023 SPIE.

关键词： Optical radar

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：