检索结果-内蒙古大学图书馆

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Bedrettin Cetinkaya Sinan Kalkan Emre Akbas Department of Computer Engineering METU ROMER Robotics Center Middle East Technical University Ankara Turkey

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Detecting edges in images suffers from the problems of (PI) heavy imbalance between positive and negative classes as well as (P2) label uncertainty owing to disagreement be-tween different annotators. Existing solutions address PI using class-balanced cross-entropy loss and dice loss and P2 by only predicting edges agreed upon by most annota-tors. In this paper, we propose RankED, a unified ranking-based approach that addresses both the imbalance problem (P 1) and the uncertainty problem (P2). RankED tackles these two problems with two components: One component which ranks positive pixels over negative pixels, and the second which promotes high confidence edge pixels to have more label certainty. We show that RankED outperforms previous studies and sets a new state-of-the-art on NYUD-v2, BSDS500 and Multi-cue datasets. Code is available at https://***.

关键词： Training computer vision Uncertainty Codes Image edge detection Detectors Pattern recognition

来源：评论

学校读者我要写书评

暂无评论

RankED: Addressing Imbalance and Uncertainty in Edge Detection Using Ranking-based Losses

arXiv

引用

arXiv 2024年

作者： Cetinkaya, Bedrettin Kalkan, Sinan Akbas, Emre Department of Computer Engineering and METU ROMER Robotics Center Middle East Technical University Ankara Turkey

Detecting edges in images suffers from the problems of (P1) heavy imbalance between positive and negative classes as well as (P2) label uncertainty owing to disagreement between different annotators. Existing solutions address P1 using class-balanced cross-entropy loss and dice loss and P2 by only predicting edges agreed upon by most annotators. In this paper, we propose RankED, a unified ranking-based approach that addresses both the imbalance problem (P1) and the uncertainty problem (P2). RankED tackles these two problems with two components: One component which ranks positive pixels over negative pixels, and the second which promotes high confidence edge pixels to have more label certainty. We show that RankED outperforms previous studies and sets a new state-of-the-art on NYUDv2, BSDS500 and Multi-cue datasets. Code is available at https://***. © 2024, CC BY.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

DeTrack: in-model latent denoising learning for visual object tracking 24

DeTrack: in-model latent denoising learning for visual objec...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Xinyu Zhou Jinglun Li Lingyi Hong Kaixun Jiang Pinxue Guo Weifeng Ge Wenqiang Zhang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Shanghai Engineering Research Center of AI & Robotics Academy for Engineering and Technology Fudan University Shanghai China Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China and Shanghai Engineering Research Center of AI & Robotics Academy for Engineering and Technology Fudan University Shanghai China

ISBN: (纸本)9798331314385

Previous visual object tracking methods employ image-feature regression models or coordinate autoregression models for bounding box prediction. Image-feature regression methods heavily depend on matching results and do not utilize positional prior, while the autoregressive approach can only be trained using bounding boxes available in the training set, potentially resulting in suboptimal performance during testing with unseen data. Inspired by the diffusion model, denoising learning enhances the model's robustness to unseen data. Therefore, We introduce noise to bounding boxes, generating noisy boxes for training, thus enhancing model robustness on testing data. We propose a new paradigm to formulate the visual object tracking problem as a denoising learning process. However, tracking algorithms are usually asked to run in real-time, directly applying the diffusion model to object tracking would severely impair tracking speed. Therefore, we decompose the denoising learning process into every denoising block within a model, not by running the model multiple times, and thus we summarize the proposed paradigm as an in-model latent denoising learning process. Specifically, we propose a denoising Vision Transformer (ViT), which is composed of multiple denoising blocks. In the denoising block, template and search embeddings are projected into every denoising block as conditions. A denoising block is responsible for removing the noise in a predicted bounding box, and multiple stacked denoising blocks cooperate to accomplish the whole denoising process. Subsequently, we utilize image features and trajectory information to refine the denoised bounding box. Besides, we also utilize trajectory memory and visual memory to improve tracking stability. Experimental results validate the effectiveness of our approach, achieving competitive performance on several challenging datasets. The proposed in-model latent denoising tracker achieve real-time speed, rendering denoising learning

关键词：

来源：评论

学校读者我要写书评

暂无评论

Lightweight Design and Property Analysis of Humanoid Robot Thigh Integrated Structure with Appearance 16th

Lightweight Design and Property Analysis of Humanoid Robot T...

引用

16th International Conference on Intelligent robotics and Applications, ICIRA 2023

作者： Nie, Daming Gu, Jason Zhang, Yu Jiang, Hongjian Research Center for Intelligent Robotics Zhejiang Lab Hangzhou310058 China Department of Electrical and Computer Engineering Dalhousie University HalifaxB3M 1A2 Canada

ISBN: (纸本)9789819964949

The lightweight and high stiffness leg structure of humanoid robots can effectively reduce rotational inertia and energy consumption, improving the robot's ability to quickly switch motion states. Based on the lightweight requirements of robots, this article proposes an innovative design method for humanoid robot legs based on the "structure-appearance" integration of light-alloy additive manufacturing technology. The load-bearing part of the leg structure is designed as a "shell + density-variable lattice" configuration, and the appearance part is set as a thin shell with equal thickness. The load-bearing part and the appearance one are combined together, resulting in a 20% weight reduction compared to traditional legs. The leg structure sample was processed using metal additive manufacturing method, and tested by a set of personalized fixture. When applying 200 Nm loading under pure torsion conditions, the local strain did not exceed 0.0175%, indicating that it meets the usage requirements. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.

关键词： Anthropomorphic robots

来源：评论

学校读者我要写书评

暂无评论

Incremental Learning with Pseudo Labels for Enhanced Glomeruli Segmentation Using SAM

Incremental Learning with Pseudo Labels for Enhanced Glomeru...

引用

2024 IEEE International Conference on E-Health Networking, Application and Services, HealthCom 2024

作者： Liu, Xinkai Peng, Boyuan Zhu, Xin Iwasaki, Tsuyoshi Sato, Atsuya Kazama, Junichiro James Huang, Jie Graduate Department of Computer Science and Engineering The University of Aizu Fukushima Aizu-Wakamatsu Japan Department of AI Technology Development M&D Data Science Center Institute of Science Tokyo Tokyo Japan Department of Nephrology and Hypertension Fukushima Medical University Fukushima Japan Graduate Department of Computer Science and Engineering Division of Information Systems The University of Aizu Fukushima Aizu-Wakamatsu Japan

ISBN: (纸本)9798350350548

The advent of the Segment Anything Model (SAM) marks a significant milestone in deep learning semantic segmentation tasks. In this study, an incremental learning model based on pseudo labels is proposed. The model generates bounding box prompts for SAM, enhancing the accuracy of glomeruli segmentation in renal pathology images. The model uses DenseNet-121 along with contrastive learning to enable self-supervised pre-training using mouse glomeruli images. Transfer learning is then conducted using labeled human renal pathology images to develop a binary classification model for glomeruli. Unlabeled images of human renal pathology are subsequently introduced for incremental learning. Class activation maps (CAMs) generated by the incremental learning model are post-processed to extract bounding rectangles for connected regions. These rectangles are then used as prompts for SAM segmentation. The proposed approach resulted in an average Dice similarity coefficient (DSC) of 0.825 in the glomeruli segmentation task. This result demonstrates the effectiveness of the incremental learning model and post-processing algorithms in improving SAM’s segmentation accuracy for renal pathology. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Optimizing Healthcare Delivery through CloudBased Clinical Decision Support Systems

Optimizing Healthcare Delivery through CloudBased Clinical D...

引用

2024 OPJU International Technology Conference on Smart Computing for Innovation and Advancement in Industry 4.0, OTCON 2024

作者： Ebenezar, U. Samson Jenila Vennila, G. Balakrishnan, T. Suresh Krishnan, Prabhakar Saveetha Institute of Medical and Technical Sciences Saveetha University Saveetha School of Engineering Department of Computer Science and Engineering Tamil Nadu Chennai India Care College of Engineering Department of Computer Science and Engineering Tamil Nadu Trichy India Mohan Babu University School of Computing Department of Ai & Ml Andhra Pradesh Tirupati India Saveetha Institute of Medical and Technical Sciences Saveetha University Saveetha School of Engineering Department of Computer Science and Engineering Chennai India Amrita Vishwa Vidyapeetham Center for Cyber Security Systems and Networks Amritapuri-campus India

ISBN: (纸本)9798350373783

The experiment aims at evaluating how four machine learning algorithms work: Decision Tree, Random Forest, SVM, and Neural Network, in a task, of binary classification. Models were evaluated using comprehensive metrics such as accuracy, precision, recall, F-score and AUC that revealed a good diagnostic performance. Models selected for the experiment and parameters tested on real datasets show steady changes in performance. The Neural Network model remained on top, achieving accuracy, precision, recall, F-score and AUC-ROC of 91.8%, 92.1%, 91.3%, 91.7% and 0.93 respectively. Random Forest displayed a competitive accordance along with the accuracy of 88.6%, precision of 89.2%, recall of 88.7%, and the F-score of 89.0%, and the AUC of 0.91. From the results, Support Vector Machine and Decision Tree models posted lower accuracy, recall, and fscores compared to the Neural Network and Random Forest models in most cases. These reveal the need for the improvement of the suitable machine learning model that comes from specific task needs as well as the consideration of each one of factors including accuracy, interpretability, and computational efficiency. Extensive and development of the model may result in more comprehensive understanding of how the model leads classifying of real-world applications such as medical diagnoses, fraud detection and target marketing. © 2024 IEEE.

关键词： Random forests

来源：评论

学校读者我要写书评

暂无评论

MoCap-to-Visual Domain Adaptation for Efficient Human Mesh Estimation from 2D Keypoints

arXiv

引用

arXiv 2024年

作者： Uguz, Bedirhan Suat, Ozhan Karagoz, Batuhan Akbas, Emre Department of Computer Engineering METU ROMER Robotics Center Middle East Technical University Ankara Turkey

This paper presents Key2Mesh, a model that takes a set of 2D human pose keypoints as input and estimates the corresponding body mesh. Since this process does not involve any visual (i.e. RGB image) data, the model can be trained on large-scale motion capture (MoCap) datasets, thereby overcoming the scarcity of image datasets with 3D labels. To enable the model’s application on RGB images, we first run an off-the-shelf 2D pose estimator to obtain the 2D keypoints, and then feed these 2D keypoints to Key2Mesh. To improve the performance of our model on RGB images, we apply an adversarial domain adaptation (DA) method to bridge the gap between the MoCap and visual domains. Crucially, our DA method does not require 3D labels for visual data, which enables adaptation to target sets without the need for costly labels. We evaluate Key2Mesh for the task of estimating 3D human meshes from 2D keypoints, in the absence of RGB and mesh label pairs. Our results on widely used H3.6M and 3DPW datasets show that Key2Mesh sets the new state-of-the-art by outperforming other models in PA-MPJPE for both datasets, and in MPJPE and PVE for the 3DPW dataset. Thanks to our model’s simple architecture, it operates at least 12× faster than the prior state-of-the-art model, LGD [44]. Additional qualitative samples and code are available on the project website: https://***/. © 2024, CC BY.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Broad Federated Meta-Learning of Damaged Objects in Aerial Videos

引用

computer Modeling in engineering & Sciences 2023年第12期137卷 2881-2899页

作者： Zekai Li Wenfeng Wang School of Aerospace Engineering Xi’an Jiaotong UniversityXi’an710049China Research Institute of Intelligent Engineering and Data Applications Shanghai Institute of TechnologyShanghai201418China Research Center for Ecology and Environment of Central Asia Chinese Academy of SciencesUrumqi830011China Applied Nonlinear Science Lab Anand International College of EngineeringJaipur391320India Department of Visual Engineering International Academy of Visual Art and EngineeringLondonCR26EQUK Institute of Intelligent Management and Technology Sino-Indian Joint Research Center of AI and RoboticsBhubaneswar752054India

We advanced an emerging federated learning technology in city intelligentization for tackling a real challenge-to learn damaged objects in aerial ***-learning system was integrated with the fuzzy broad learning system to further develop the theory of federated *** the mixed picture set of aerial video segmentation and the 3D-reconstructed mixed-reality data were employed in the performance of the broad federated meta-learning *** study results indicated that the object classification accuracy is up to 90%and the average time cost in damage detection is only 0.277 ***,the broad federated meta-learning system is efficient and effective in detecting damaged objects in aerial videos.

关键词： Fuzzy learning system mixed-reality 3D-reconstructed oblique photography

来源：评论

学校读者我要写书评

暂无评论

Surgical Tool Occlusion Detection and Segmentation Using Deep Transfer Learning: A Machine Vision Approach

Surgical Tool Occlusion Detection and Segmentation Using Dee...

引用

2024 International Conference on Decision aid Sciences and Applications, DASA 2024

作者： Sevilla, Rovenson Alon, Alvin Sarraga Evangelista, Ryan S. Delos Santos Ligayo, Michael Angelo Arago, Ryan Flores Delos Santos, Ronnel Cabral Cunanan, Christopher Austria, Yolanda D. University of the Philippines Department of Electrical Engineering Technological Manila Philippines Center for AI Research Department of Trade and Industry Makati City Philippines School of Graduate Studies Sulu State College Sulu Philippines College of Engineering Quezon City University Quezon City Philippines College of Computing Studies Don Honorio Ventura State University Pampanga Philippines Bulacan State University Computer Engineering Department Bulacan Philippines Adamson University Department of Computer Engineering Manila Philippines

ISBN: (纸本)9798350369106

The combination of deep transfer learning and mask region-based convolutional neural network (Mask-RCNN) technology offers a groundbreaking approach to enhancing machine vision for surgical robots. The objective of this work is to enhance the accuracy and precision of identifying tools in surgical environments by investigating the use of deep transfer learning for detecting and segmenting occluded surgical tools. Our investigation exhibits the model's robustness in successfully recognizing and segmenting obstructed surgical tools, as seen by an exceptional Mean Average Precision (mAP) score of 93.8%. By conducting thorough testing and inference studies, the model demonstrated an outstanding ability to accurately identify various instruments, such as 'mayo,' 'clamp,' and 'scalpel,' even in difficult situations when objects were partially occluded. The use of Mask-RCNN enhances our investigation, representing a crucial advancement in enhancing surgical robotics with sophisticated segmentation skills. The research underscores the potential of Mask-RCNN in the field of machine vision for surgical robots, underlining the necessity for ongoing progress to tackle complexities for improved real-world implementation. © 2024 IEEE.

关键词： Machine vision

来源：评论

学校读者我要写书评

暂无评论

PG-Attack: A Precision-Guided Adversarial Attack Framework Against Vision Foundation Models for Autonomous Driving

arXiv

引用

arXiv 2024年

作者： Fu, Jiyuan Chen, Zhaoyu Jiang, Kaixun Guo, Haijing Gao, Shuyong Zhang, Wenqiang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University China Shanghai Engineering Research Center of AI & Robotics Academy for Engineering & Technology Fudan University China Engineering Research Center of Robotics Ministry of Education Academy for Engineering & Technology Fudan University China

Vision foundation models are increasingly employed in autonomous driving systems due to their advanced capabilities. However, these models are susceptible to adversarial attacks, posing significant risks to the reliability and safety of autonomous vehicles. Adversaries can exploit these vulnerabilities to manipulate the vehicle’s perception of its surroundings, leading to erroneous decisions and potentially catastrophic consequences. To address this challenge, we propose a novel Precision-Guided Adversarial Attack (PG-Attack) framework that combines two techniques: Precision Mask Perturbation Attack (PMP-Attack) and Deceptive Text Patch Attack (DTP-Attack). PMP-Attack precisely targets the attack region to minimize the overall perturbation while maximizing its impact on the target object’s representation in the model’s feature space. DTP-Attack introduces deceptive text patches that disrupt the model’s understanding of the scene, further enhancing the attack’s effectiveness. Our experiments demonstrate that PG-Attack successfully deceives a variety of advanced multi-modal large models, including GPT-4V, Qwen-VL, and imp-V1. Additionally, we won First-Place in the CVPR 2024 Workshop Challenge: Black-box Adversarial Attacks on Vision Foundation Models‡ and codes are available at https://***/fuhaha824/PG-Attack. Copyright © 2024, The Authors. All rights reserved.

关键词： Autonomous vehicles

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：