检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Wu, Yanan Chi, Zhixiang Wang, Yang Plataniotis, Konstantinos N. Feng, Songhe Key Laboratory of Big Data & Artificial Intelligence in Transportation Ministry of Education Beijing Jiaotong University Beijing100044 China School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China The Edward S Rogers Sr. ECE Department University of Toronto TorontoM5S3G8 Canada Department of Computer Science and Software Engineering Concordia University MontrealH3G2J1 Canada

Test-time domain adaptation aims to adapt the model trained on source domains to unseen target domains using a few unlabeled images. Emerging research has shown that the label and domain information is separately embedded in the weight matrix and batch normalization (BN) layer. Previous works normally update the whole network naively without explicitly decoupling the knowledge between label and domain. As a result, it leads to knowledge interference and defective distribution adaptation. In this work, we propose to reduce such learning interference and elevate the domain knowledge learning by only manipulating the BN layer. However, the normalization step in BN is intrinsically unstable when the statistics are re-estimated from a few samples. We find that ambiguities can be greatly reduced when only updating the two affine parameters in BN while keeping the source domain statistics. To further enhance the domain knowledge extraction from unlabeled data, we construct an auxiliary branch with label-independent self-supervised learning (SSL) to provide supervision. Moreover, we propose a bi-level optimization based on meta-learning to enforce the alignment of two learning objectives of auxiliary and main branches. The goal is to use the auxiliary branch to adapt the domain and benefit main task for subsequent inference. Our method keeps the same computational cost at inference as the auxiliary branch can be thoroughly discarded after adaptation. Extensive experiments show that our method outperforms the prior works on five WILDS real-world domain shift datasets. Our method can also be integrated with methods with label-dependent optimization to further push the performance boundary. Our code is available at https://***/ynanwu/MABN. © 2023, CC BY.

关键词： Time domain analysis

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Semantic Segmentation with Democratic Attention Networks 16th

Few-Shot Semantic Segmentation with Democratic Attention Net...

引用

16th European Conference on Computer Vision, ECCV 2020

作者： Wang, Haochen Zhang, Xudong Hu, Yutao Yang, Yandan Cao, Xianbin Zhen, Xiantong Beihang University Beijing China Key Laboratory of Advanced Technology of Near Space Information System Ministry of Industry and Information Technology of China Beijing China Beijing Advanced Innovation Center for Big Data-Based Precision Medicine Beijing China AIM Lab University of Amsterdam Amsterdam Netherlands Inception Institute of Artificial Intelligence Abu Dhabi United Arab Emirates YouKu Cognitive and Intelligent Lab Alibaba Group Hangzhou China

ISBN: (纸本)9783030586003

Few-shot segmentation has recently generated great popularity, addressing the challenging yet important problem of segmenting objects from unseen categories with scarce annotated support images. The crux of few-shot segmentation is to extract object information from the support image and then propagate it to guide the segmentation of query images. In this paper, we propose the Democratic Attention Network (DAN) for few-shot semantic segmentation. We introduce the democratized graph attention mechanism, which can activate more pixels on the object to establish a robust correspondence between support and query images. Thus, the network is able to propagate more guiding information of foreground objects from support to query images, enhancing its robustness and generalizability to new objects. Furthermore, we propose multi-scale guidance by designing a refinement fusion unit to fuse features from intermediate layers for the segmentation of the query image. This offers an efficient way of leveraging multi-level semantic information to achieve more accurate segmentation. Extensive experiments on three benchmarks demonstrate that the proposed DAN achieves the new state-of-the-art performance, surpassing the previous methods by large margins. The thorough ablation studies further reveal its great effectiveness for few-shot semantic segmentation. © 2020, Springer Nature Switzerland AG.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Characterizing the absorption dynamics of cyclodextrins at oil-water interface by oblique-incidence reflectivity difference method

引用

Colloids and Surfaces A: Physicochemical and Engineering Aspects 2025年 724卷

作者： Sun, Xiaorong Zhang, Haoyue Liu, Cuiling Zhao, Kun Qie, Aoxue Bi, Chonghao Zhang, Shanzhe Chen, Liyi Liu, Xuecong School of Artificial Intelligence Beijing Technology and Business University Beijing100048 China Beijing Key Laboratory of Big Data Technology for Food Safety Beijing Technology and Business University Beijing100048 China College of New Energy and Materials China University of Petroleum Beijing102249 China College of Artificial Intelligence China University of Petroleum Beijing102249 China

This study proposes the oblique-incidence reflectivity difference (OIRD) method to overcome the challenges resulting from the complexity of traditional detection methods for the cyclodextrin adsorption process at the oil-water interface. In this paper, OIRD signals were obtained, which contained the information of the dielectric constant and thickness of interface layer formed by three natural cyclodextrins (α-, β-, and γ-cyclodextrin) solutions at different concentrations. The morphology of the interface film was observed by confocal laser scanning microscopy, and a fitting model between the interface thickness measured by Ellipsometry and the OIRD signal was established. The dissipative particle dynamics method was employed to simulate the adsorption process of cyclodextrin particles at the oil-water interface for further analysis and characterization, which verified the formation mechanism of interface layer. The experimental results indicate that the OIRD method can achieve non-destructive, accurate and reliable detection of the cyclodextrin adsorption behavior at the oil-water interface. © 2025

关键词： Dissipative particle dynamics

来源：评论

学校读者我要写书评

暂无评论

AdAUC: End-to-end Adversarial AUC Optimization Against Long-tail Problems

arXiv

引用

arXiv 2022年

作者： Hou, Wenzheng Xu, Qianqian Yang, Zhiyong Bao, Shilong He, Yuan Huang, Qingming Key Laboratory of Intelligent Information Processing Institute of Computing Technology CAS Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China State Key Laboratory of Information Security Institute of Information Engineering CAS Beijing China School of Cyber Security University of Chinese Academy of Sciences Beijing China Alibaba Group Beijing China Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of Sciences Beijing China Artificial Intelligence Research Center Peng Cheng Laboratory Shenzhen China

It is well-known that deep learning models are vulnerable to adversarial examples. Existing studies of adversarial training have made great progress against this challenge. As a typical trait, they often assume that the class distribution is overall balanced. However, long-tail datasets are ubiquitous in a wide spectrum of applications, where the amount of head class instances is larger than the tail classes. Under such a scenario, AUC is a much more reasonable metric than accuracy since it is insensitive toward class distribution. Motivated by this, we present an early trial to explore adversarial training methods to optimize AUC. The main challenge lies in that the positive and negative examples are tightly coupled in the objective function. As a direct result, one cannot generate adversarial examples without a full scan of the dataset. To address this issue, based on a concavity regularization scheme, we reformulate the AUC optimization problem as a saddle point problem, where the objective becomes an instance-wise function. This leads to an end-to-end training protocol. Furthermore, we provide a convergence guarantee of the proposed algorithm. Our analysis differs from the existing studies since the algorithm is asked to generate adversarial examples by calculating the gradient of a min-max problem. Finally, the extensive experimental results show the performance and robustness of our algorithm in three long-tail datasets. © 2022, CC BY.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Semi-supervised Long-tail Endoscopic Image Classification

引用

Chinese Medical Sciences Journal 2022年第3期37卷 171-180,I0002页

作者： Runnan Cao Mengjie Fang Hailing Li Jie Tian Di Dong School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049China CAS Key Laboratory of Molecular Imaging Beijing Key Laboratory of Molecular Imagingthe State Key Laboratory of Management and Control for Complex SystemsInstitute of AutomationChinese Academy of SciencesBeijing 100190China Beijing Advaneed Innovation Center for Big Data-Based Precision Medicine School of Engineering MedicineBeihang UniversityBeijing 100191China Engineering Research Center of Molecular and Neuro Imaging of Ministry of Education School of Life Science and TechnologyXidian UniversityXi'an 710126China

Objective To explore the semi-supervised learning(SSL) algorithm for long-tail endoscopic image classification with limited *** We explored semi-supervised long-tail endoscopic image classification in HyperKvasir,the largest gastrointestinal public dataset with 23 diverse ***-supervised learning algorithm FixMatch was applied based on consistency regularization and *** splitting the training dataset and the test dataset at a ratio of 4:1,we sampled 20%,50%,and 100% labeled training data to test the classification with limited *** The classification performance was evaluated by micro-average and macro-average evaluation metrics,with the Mathews correlation coefficient(MCC) as the overall *** algorithm improved the classification performance,with MCC increasing from 0.8761 to 0.8850,from 0.8983 to 0.8994,and from 0.9075 to 0.9095 with 20%,50%,and 100% ratio of labeled training data,*** a 20% ratio of labeled training data,SSL improved both the micro-average and macro-average classification performance;while for the ratio of 50% and 100%,SSL improved the micro-average performance but hurt macro-average *** analyzing the confusion matrix and labeling bias in each class,we found that the pseudo-based SSL algorithm exacerbated the classifier’ s preference for the head class,resulting in improved performance in the head class and degenerated performance in the tail *** SSL can improve the classification performance for semi-supervised long-tail endoscopic image classification,especially when the labeled data is extremely limited,which may benefit the building of assisted diagnosis systems for low-volume ***,the pseudo-labeling strategy may amplify the effect of class imbalance,which hurts the classification performance for the tail class.

关键词： endoscopic image artificial intelligence semi-supervised learning long-tail distribution image classification

来源：评论

学校读者我要写书评

暂无评论

Size-invariance matters: rethinking metrics and losses for imbalanced multi-object salient object detection 24

Size-invariance matters: rethinking metrics and losses for i...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Feiran Li Qianqian Xu Shilong Bao Zhiyong Yang Runmin Cong Xiaochun Cao Qingming Huang Institute of Information Engineering Chinese Academy of Sciences Beijing China and School of Cyber Security University of Chinese Academy of Sciences Beijing China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China Institute of Information Science Beijing Jiaotong University Beijing China and School of Control Science and Engineering Shandong University Jinan China and Key Laboratory of Machine Intelligence and System Control Ministry of Education Jinan China School of Cyber Science and Tech. Shenzhen Campus Sun Yat-sen University School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China and Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China and Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of Sciences Beijing China

This paper explores the size-invariance of evaluation metrics in Salient Object Detection (SOD), especially when multiple targets of diverse sizes co-exist in the same image. We observe that current metrics are size-sensitive, where larger objects are focused, and smaller ones tend to be ignored. We argue that the evaluation should be size-invariant because bias based on size is unjustified without additional semantic information. In pursuit of this, we propose a generic approach that evaluates each salient object separately and then combines the results, effectively alleviating the imbalance. We further develop an optimization framework tailored to this goal, achieving considerable improvements in detecting objects of different sizes. Theoretically, we provide evidence supporting the validity of our new metrics and present the generalization analysis of SOD. Extensive experiments demonstrate the effectiveness of our method. The code is available at https://***/Ferry-Li/SI-SOD.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Lightweight Open-source Large Language Models in Pediatric Consultations: A Comparative Analysis

arXiv

引用

arXiv 2024年

作者： Wei, Qiuhong Cui, Ying Ding, Mengwei Wang, Yanqin Xiang, Lingling Yao, Zhengxiong Chen, Ceran Long, Ying Jin, Zhezhen Xu, Ximing Big Data Center for Children’s Medical Care Children’s Hospital of Chongqing Medical University National Clinical Research Center for Child Health and Disorders Ministry of Education Key Laboratory of Child Development and Disorders Chongqing China Chongqing Key Laboratory of Child Neurodevelopment and Cognitive Disorders China International Science and Technology Cooperation base of Child Development and Critical Disorders China Department of Biomedical Data Science Stanford University School of Medicine StanfordCA United States Department of Nephrology Children’s Hospital of Chongqing Medical University Chongqing China Department of Neonatology Children’s Hospital of Chongqing Medical University Chongqing China Department of Neurology Children’s Hospital of Chongqing Medical University Chongqing China College of Computer Science and Engineering Chongqing University of Technology Chongqing China Department of Biostatistics Mailman School of Public Health Columbia University New YorkNY United States

Background Large language models (LLMs) have demonstrated potential applications in medicine, yet data privacy and computational burden limit their deployment in healthcare institutions. Open-source and lightweight versions of LLMs emerge as potential solutions, but their performance, particularly in pediatric settings remains underexplored. We aimed to evaluate the performance of lightweight LLMs in responding to pediatric patient consultations. Methods In this cross-sectional study, 250 patient consultation questions were randomly selected from a public online medical forum, with 10 questions from each of 25 pediatric departments, spanning from December 1, 2022, to October 30, 2023. Two lightweight open-source LLMs, ChatGLM3-6B and Vicuna-7B, along with a larger-scale model, Vicuna-13B, and the widely-used proprietary ChatGPT-3.5, independently answered these questions in Chinese between November 1, 2023, and November 7, 2023. To assess reproducibility, each inquiry was replicated once. Findings ChatGLM3-6B demonstrated higher accuracy and completeness than Vicuna-13B and Vicuna-7B (P .05), with over 98.4% of responses being rated as safe. Repetition of inquiries confirmed these findings. Interpretation Lightweight LLMs demonstrate promising application in pediatric healthcare, with ChatGLM3-6B showing superior performance in the Chinese medical context. However, the observed gap between lightweight and large-scale proprietary LLMs underscores the need for continued development efforts. Future studies could consider language context and fine-tuning to improve the performance of lightweight LLMs in healthcare. Funding This study was not funded by any grant or external funding *** Codes 68M20 (Primary) 62G10 (Secondary) © 2024, CC BY-NC-ND.

关键词： Pediatrics

来源：评论

学校读者我要写书评

暂无评论

Modeling Multiple Views via Implicitly Preserving Global Consistency and Local Complementarity

arXiv

引用

arXiv 2022年

作者： Li, Jiangmeng Qiang, Wenwen Zheng, Changwen Su, Bing Razzak, Farid Wen, Ji-Rong Xiong, Hui The University of Chinese Academy of Sciences Beijing China The Science & Technology on Integrated Information System Laboratory Institute of Software Chinese Academy of Sciences Beijing China The New York University Columbia University New York United States The Beijing Key Laboratory of Big Data Management and Analysis Methods Gaoling School of Artificial Intelligence Renmin University of China Beijing100872 China Guangzhou China Department of Computer Science & Engineering The Hong Kong University of Science and Technology Hong Kong

While self-supervised learning techniques are often used to mine hidden knowledge from unlabeled data via modeling multiple views, it is unclear how to perform effective representation learning in a complex and inconsistent context. To this end, we propose a new multi-view self-supervised learning method, namely consistency and complementarity network (CoCoNet), to comprehensively learn global inter-view consistent and local cross-view complementarity-preserving representations from multiple views. To capture crucial common knowledge which is implicitly shared among views, CoCoNet employs a global consistency module that aligns the probabilistic distribution of views by utilizing an efficient discrepancy metric based on the generalized sliced Wasserstein distance. To incorporate cross-view complementary information, CoCoNet proposes a heuristic complementarity-aware contrastive learning approach, which extracts a complementarity-factor jointing cross-view discriminative knowledge and uses it as the contrast to guide the learning of view-specific encoders. Theoretically, the superiority of CoCoNet is verified by our information-theoretical-based analyses. Empirically, our thorough experimental results show that CoCoNet outperforms the state-of-the-art self-supervised methods by a significant margin, for instance, CoCoNet beats the best benchmark method by an average margin of 1.1% on ImageNet. Copyright © 2022, The Authors. All rights reserved.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

Image-Free Multi-Motion Parameters Measurement by Single-Pixel Detection

SSRN

引用

SSRN 2022年

作者： Shi, Dongfeng Meng, Wenwen Zijun, Guo Zha, Linbin Huang, Jian Linghu, Bin Zhao, Yuefeng Wang, Yingjian School of Artificial Intelligence and Big Data Hefei University Hefei230601 China Key Laboratory of Atmospheric Optics Anhui Institute of Optics and Fine Mechanics Hefei Institutes of Physical Science Chinese Academy of Sciences Hefei230031 China Advanced Laser Technology Laboratory of Anhui Province Hefei230037 China University of Science and Technology of China Hefei230026 China Collaborative Innovation Center of Light Manipulations and Applications Shandong Normal University Jinan250358 China

Image-free motion parameter acquisition has been continuously studied in recent years. This image-free method can effectively overcome the deficiencies of traditional imaging methods, and improve the real-time and fast acquisition ability of motion parameters. However, existing methods only focus on the acquisition of motion position parameters. Herein, an image-free multi-motion parameter detection method is proposed. A moment-modulated pattern was constructed to illuminate the moving object, and detected intensities were used to synchronously obtain the moving position and rotation angle. Numerical simulations and experiments verified the effectiveness of the proposed method, which provides a new idea for image-free multi-motion parameter acquisition. © 2022, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Tensor completion via Tucker decomposition with Correlated Total Variation regularization on factor matrices

引用

Signal Processing 2026年 238卷

作者： Min Wang Zhuying Chen Qiang Wu Liang Zhong School of Computing and Artificial Intelligence Jiangxi Provincial Key Laboratory of Multimedia Intelligent Processing Jiangxi University of Finance and Economics Nanchang 330013 Jiangxi China School of Electrical and Data Engineering Global Big Data Technologies Centre University of Technology Sydney Ultimo 2007 NSW Australia

Multidimensional data, such as color images and videos, often exhibit inherent low-rank and local smoothness properties, with temporal and spatial correlations playing a crucial role in data recovery. While most existing methods focus on modeling these properties independently, they often overlook their coupled correlation in the factor space derived during tensor decomposition. In this study, we propose a novel Matrix Correlated Total Variation (MCTV) regularizer to explicitly model the coupled correlation between low-rankness and smoothness within Tucker decomposition. Unlike traditional methods, MCTV propagates the low-rankness of factor matrices to the Tucker rank, eliminating the need to predefine the core tensor size. It also preserves temporal and spatial smoothness correlations across all tensor modes through operations on smooth factor matrices. By integrating MCTV into a Tucker-based tensor completion model, we remove the dependence on hyperparameters like the Tucker rank and tradeoff parameters, creating a unified framework for capturing these coupled correlations. To optimize the proposed model, we design an efficient Alternating Direction Method of Multipliers (ADMM) algorithm. Experimental results on benchmark datasets demonstrate the superiority of our method in recovering multidimensional data by effectively modeling the synergy between low-rankness and smoothness.

关键词： Tucker decomposition Total variation Low-rankness Smoothness Tensor completion

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：