检索结果-内蒙古大学图书馆

29th International Conference on Neural Information Processing

作者： Zhao, Ming Wei, Xing Lu, Yang Bai, Ting Zhao, Chong Chen, Lei Hu, Di Hefei Univ Technol Sch Comp & Informat Hefei Peoples R China Hefei Univ Technol Intelligent Mfg Technol Res Inst Hefei Peoples R China Hefei Univ Technol Intelligent Interconnected Syst Lab Anhui Prov Hefei Peoples R China Chinese Acad Sci Inst Intelligent Machines HFIPS Hefei Peoples R China

ISBN: (纸本)9789819916443;9789819916450

cross-domain object detection usually solves the problem of domain transfer by reducing the difference between the source domain and target domain. However, existing solutions do not effectively solve the performance degradation caused by cross-domain differences. To address this problem, we present the cross-domain object detection Model via Contrastive Learning with Style Transfer(COCS). Our model is based on generating new samples with source domain information and target domain style. In addition, the importance of new samples feature information are aimed to match positive and negative samples for comparative learning better. So, we transfer source domain with labeled to get new samples with style of target domain. Then we employ momentum contrast learning method to maximize the similarly between positive sample pairs representations and minimize the loss function. Moreover, our model can be adapted to different style domains, which further expands the application scenarios. Experiments on a benchmark dataset demonstrate that our model achieves or matches the state-of-the-art approaches.

关键词： cross-domain object detection Style Transfer Contrastive learning

来源：评论

学校读者我要写书评

暂无评论

Unsupervised cross-domain object detection based on dynamic smooth cross entropy

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2025年 1-14页

作者： Xie, Bojun Huang, Zhijin Chen, Junfen Key Lab Machine Learning & Computat Intelligence Baoding 071002 Hebei Peoples R China Hebei Univ Coll Math & Informat Sci Baoding 071002 Hebei Peoples R China

Pseudo-label self-training methods are highly regarded in cross-domain object detection (CDOD) models for their reduced annotation costs, strong applicability, and straightforward deployment. However, existing methods still struggle to balance the quantity and quality of pseudo-labels, as well as their efficient utilization. To tackle these challenges, this study introduces a novel approach by integrating strong-weak data augmentation (SWDA) techniques into the decoupled adaptation framework, significantly enhancing the generalization ability and accuracy of the pseudo-label generator. Furthermore, inspired by Softmatch, this research proposes a dynamic smooth cross-entropy loss function that adjusts pseudo-label confidence thresholds in real-time based on the model's learning state and the predictive distribution across categories. This method not only retains low-confidence pseudo-labels but also assigns varied weights to them according to their confidence levels, enhancing both the quality and utility of pseudo-labels. This approach substantially improves CDOD performance, especially demonstrated in the Pascal VOC ->\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\rightarrow$$\end{document} Clipart task, where it outperformed the leading Adaptive Teacher and Class-Aware Teacher methods by 2.3 and 2.5% points, respectively, achieving an accuracy of 51.6%. The method also demonstrated commendable performance in other commonly used unsupervised CDOD tasks. The code is available at https://***/985767019/DSCE.

关键词： Decoupled adaptation Strong-weak data augmentation Dynamic smooth cross entropy Pseudo-Labels cross-domain object detection

来源：评论

学校读者我要写书评

暂无评论

THNet: Transferability-Aware Hierarchical Network for Robust cross-domain object detection

引用

IEEE ACCESS 2024年 12卷 155469-155484页

作者： Song, Wu Ren, Sheng Tan, Wenxue Wang, Xiping Hunan Univ Arts & Sci Sch Comp & Elect Engn Changde 415000 Peoples R China

Deep learning has advanced object detection, but generalizing models from source to target domains remains a challenge due to multi-level domain drift and untransferable information. To address this, we propose a transferability-aware hierarchical domain-consistent object detector (THNet), incorporating instance-level, pixel-level, and image-level alignment subnets for robust cross-domain detection. THNet first aligns local foreground-transferable features through pixel-level adversarial learning and foreground-aware attention, then captures global domain-invariant features via image-level subnet with channel-transferable attention. Additionally, a prototype graph convolutional network alleviates instance distribution differences by maximizing inter-class distances and minimizing intra-class distances. A domain-consistent loss harmonizes training for better convergence in multi-level domain alignment. Extensive experiments demonstrate that THNet outperforms state-of-the-art methods on multiple cross-domain datasets, achieving top accuracies of 51.9%, 46.0%, 41.2%, and 51.9% across different tasks.

关键词： object detection Feature extraction Training Detectors Adversarial machine learning Adaptation models Prototypes Residual neural networks Remote sensing Mathematical models cross-domain object detection hierarchical domain alignment domain-consistent loss transferable attention adversarial learning

来源：评论

学校读者我要写书评

暂无评论

Dual Instance-Consistent Network for cross-domain object detection

引用

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 2023年第6期45卷 7338-7352页

作者： Jiao, Yifan Yao, Hantao Xu, Changsheng Nanjing Univ Posts & Telecommun Sch Commun & Informat Engn Nanjing Peoples R China Chinese Acad Sci Inst Automat Natl Lab Pattern Recognit Beijing Peoples R China Chinese Acad Sci Inst Automation Natl Lab Pattern Recognit Beijing Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing 100049 Peoples R China Peng Cheng Lab Shenzhen 518055 Peoples R China

cross-domain object detection aims to transfer knowledge from a labeled dataset to an unlabeled dataset. Most existing methods apply a unified embedding model to generate the tightly coupled source and target descriptions for domain alignment, leading to the destroyed feature distribution of the target domain because the embedding model is mainly controlled by the source domain. To reduce the representation bias of the target domain, we apply two independent networks to extract two types of discriminative descriptions with mutual consistency, i.e., a novel Dual Instance-Consistent Network (DICN) is proposed for cross-domain object detection. Especially, Dual Instance-Consistent Module containing the instance mutual consistency between Primary Network and Auxiliary Network is applied to align two domains, where Primary and Auxiliary Networks are used to obtain the source-specific and target-specific information, respectively. The instance mutual consistency consists of two terms: feature consistency and detection consistency, which is applied to align the instance feature and the output of detection head, respectively. With the instance mutual consistency, optimizing the Primary (Auxiliary) Network only with source (target) images by fixing the Auxiliary (Primary) Network can generate the source(target)-specific description. Extensive experiments on several benchmarks demonstrate the effectiveness of the proposed DICN, e.g., obtaining mAP of 44.10% for Cityscapes-> Foggy Cityscapes, AP on car of 76.50% for Cityscapes-> KITTI, MR (2) of 8.87%, 12.66%, 22.27%, and 42.06% for COCOPersons-> Caltech, CityPersons-> Caltech, COCOPersons-> CityPersons, and Caltech-> CityPersons, respectively.

关键词： Feature extraction object detection Detectors Visualization Proposals Head Task analysis cross-domain object detection domain-specific description dual instance-consistent network

来源：评论

学校读者我要写书评

暂无评论

Teacher-Student cross-domain object detection Model Combining Style Transfer and Adversarial Learning 6th

Teacher-Student Cross-Domain Object Detection Model Combinin...

引用

6th Chinese Conference on Pattern Recognition and Computer Vision (PRCV)

作者： Wu, Lijun Cao, Zhe Chen, Zhicong Fuzhou Univ Sch Adv Mfg Fujian Peoples R China

ISBN: (纸本)9789819985487;9789819985494

cross-domain object detection is challenging because object detection models are significantly susceptible to domain style. As a popular semi-supervised learning method, the teacher-student framework (pseudo labels from the teacher model supervise the student model) achieves significant accuracy gains in cross-domain object detection. However, it suffers from the domain shift and prone to generate low-quality pseudo labels, which limits the performance. To mitigate this problem, we propose a teacher-student framework that utilizes style transfer method, augmentation strategies, and adversarial learning to address domain shift. Specifically, we design a Fourier style transfer method to reduce the gap between source and target domainswithout altering the semantic information of the objects. Furthermore, we improve the data augmentation strategy, by weakly augmenting the images from the target domain, to avoid the teacher model biased to the source domain. Finally, we employ feature-level adversarial training in the student model which is trained based on images from all domains, allowing features derived from all domains to share similar distributions. This process ensures that the student model produces domain-invariant features. Our approach achieves state-of-the-art performance in several benchmark tests. For example, it achieved 51.6% and 49.9% mAP on Foggy Cityscapes and Clipart1K, respectively.

关键词： cross-domain object detection Unsupervised domain Adaptation Style Transfer Adversarial Learning

来源：评论

学校读者我要写书评

暂无评论

DIVERGENCE-GUIDED FEATURE ALIGNMENT FOR cross-domain object detection 47

DIVERGENCE-GUIDED FEATURE ALIGNMENT FOR CROSS-DOMAIN OBJECT ...

引用

47th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

作者： Li, Zongyao Togo, Ren Ogawa, Takahiro Haseyama, Miki Hokkaido Univ Grad Sch Informat Sci & Technol Sapporo Hokkaido Japan Hokkaido Univ Educ & Res Ctr Math & Data Sci Sapporo Hokkaido Japan Hokkaido Univ Fac Informat Sci & Technol Sapporo Hokkaido Japan

ISBN: (纸本)9781665405409

domain shift causes performance drop in cross-domain object detection. To alleviate the domain shift, a prevailing approach is global feature alignment with adversarial learning. However, such simple feature alignment has defects of unawareness of foreground/background regions and well-aligned/poorly-aligned regions. To remedy the defects, in this paper, we propose a novel divergence-guided feature alignment method for cross-domain object detection. Specifically, we generate source-like images of the target domain and seek cues of foreground regions and poorly-aligned regions from prediction divergence of the source-like and original images. The feature alignment is guided by the divergence maps and consequently results in adaptation performance superior to alignment unaware of the cues. Different from most previous studies focusing on two-stage object detection, this paper is devoted to adapting one-stage object detectors which have simpler and faster inference. We validated the effectiveness of our method by conducting experiments in cross-weather, cross-camera, and synthetic-to-real adaptation scenarios.

关键词： cross-domain object detection one-stage object detection unsupervised domain adaptation

来源：评论

学校读者我要写书评

暂无评论

Sequential Instance Refinement for cross-domain object detection in Images

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2021年 30卷 3970-3984页

作者： Chen, Jin Wu, Xinxiao Duan, Lixin Chen, Lin Beijing Inst Technol Beijing Lab Intelligent Informat Technol Beijing 100081 Peoples R China Beijing Inst Technol Sch Comp Sci Beijing 100081 Peoples R China Univ Elect Sci & Technol China Sch Comp Sci & Engn Chengdu 611731 Peoples R China Wyze Labs Kirkland WA 98034 USA

cross-domain object detection in images has attracted increasing attention in the past few years, which aims at adapting the detection model learned from existing labeled images (source domain) to newly collected unlabeled ones (target domain). Existing methods usually deal with the cross-domain object detection problem through direct feature alignment between the source and target domains at the image level, the instance level (i.e., region proposals) or both. However, we have observed that directly aligning features of all object instances from the two domains often results in the problem of negative transfer, due to the existence of (1) outlier target instances that contain confusing objects not belonging to any category of the source domain and thus are hard to be captured by detectors and (2) low-relevance source instances that are considerably statistically different from target instances although their contained objects are from the same category. With this in mind, we propose a reinforcement learning based method, coined as sequential instance refinement, where two agents are learned to progressively refine both source and target instances by taking sequential actions to remove both outlier target instances and low-relevance source instances step by step. Extensive experiments on several benchmark datasets demonstrate the superior performance of our method over existing state-of-the-art baselines for cross-domain object detection.

关键词： object detection Feature extraction Detectors Reinforcement learning Proposals Task analysis Benchmark testing cross-domain object detection negative transfer reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Tri-Flow-YOLO: Counter helps to improve cross-domain object detection

引用

HELIYON 2024年第11期10卷 e32413页

作者： Wei, Jian Wang, Qinzhao Army Acad Armored Forces Beijing 100071 Peoples R China

The excellence of intelligent detection models has been widely recognized, but in terms of crossdomain scenes, they still face performance degradation and low accuracy. A multi-supervised TriFlow-YOLO model is proposed to improve the accuracy of objects with various scales under crossdomain conditions. Based on the full-supervised traditional detection branch of YOLOv5, another two mutually supporting task branches are designed intently. In brief, we add unsupervised adversarial classification training flow to the backend, to realize the feature alignment requirements and improve the cross-domain performance stability of the model. Meanwhile, a weakly-supervised object counting flow is proposed to improve the model's attention to all the objects and the detection ability is efficiently enforced. In addition, I-Mosaic and iCIOU are designed especially for small hard objects, enriching the positive samples during the training process. With the auxiliary of both improved strategies, the imbalance of positive and negative samples in the anchor-based model is relieved accordingly. The experimental results show that the improved Tri-Flow-YOLO model achieves 56.0 mAP in the Cityscapes-*Foggy-Cityscapes task, and 49.8 mAP in the VOC-*Clipart task.

关键词： cross-domain object detection Multi-task Mosaic Unsupervised domain adaptation

来源：评论

学校读者我要写书评

暂无评论

An Efficient and Accurate cross-domain object detection Method Using One-Level Feature and domain Adaptation

引用

INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE 2023年第7期37卷 2350016-2350016页

作者： Zhang, Tianyuan Song, Xudong Zhu, Chen Liang, Pan Sun, Jialiang Wang, Shuo Cui, Yunxian Li, Changxian Dalian Jiaotong Univ Big Data & Intelligent Syst Lab Dalian 116028 Liaoning Peoples R China

To improve the accuracy of cross-domain object detection, the existing unsupervised domain adaptation (UDA) object detection methods mostly use Feature Pyramid Network (FPN), multiple Region Proposal Network (RPN), and multiple domain classifier, but these methods lead to complex network structures, slow model convergence, and low detection efficiency. To solve the above problems, this paper proposes an Efficient and Accurate cross-domain object detection Method Using One-level Feature and domain Adaptation (OFDA). This method realizes one-level feature object detection through feature fusion and divide-and-conquer technology;realizes overfitting feature suppression and unsupervised domain adaptation through domain-specific suppression and domain feature alignment technology;realizes background feature suppression through the objectness branch, which replaces the time-consuming Region Proposal Network (RPN) structure and improves the efficiency of unsupervised adaptive detection. The paper verifies the feasibility and superiority of the proposed method by the comparative experiments and ablation experiments of multiple datasets. The proposed OFDA method not only improves the efficiency of object detection, but also ensures the accuracy of cross-domain detection.

关键词： cross-domain object detection unsupervised domain adaptation multi-scale one-level feature

来源：评论

学校读者我要写书评

暂无评论

HIERARCHICAL domain-CONSISTENT NETWORK FOR cross-domain object detection

HIERARCHICAL DOMAIN-CONSISTENT NETWORK FOR CROSS-DOMAIN OBJE...

引用

IEEE International Conference on Image Processing (ICIP)

作者： Liu, Yuanyuan Liu, Ziyang Fang, Fang Fu, Zhanghua Chen, Zhanlong China Univ Geosci Sch Geog & Informat Engn Wuhan Peoples R China Chinese Univ Hong Kong Inst Robot & Intelligent Mfg Shenzhen Peoples R China

ISBN: (纸本)9781665441155

cross-domain object detection is a very challenging task due to multi-level domain shift in an unseen domain. To address the problem, this paper proposes a hierarchical domain-consistent network (HDCN) for cross-domain object detection, which effectively suppresses pixel-level, image-level, as well as instance-level domain shift via jointly aligning three-level features. Firstly, at the pixel-level feature alignment stage, a pixel-level subnet with foreground-aware attention learning and pixel-level adversarial learning is proposed to focus on local foreground transferable information. Then, at the image-level feature alignment stage, global domain-invariant features are learned from the whole image through image-level adversarial learning. Finally, at the instance-level alignment stage, a prototype graph convolution network is conducted to guarantee distribution alignment of instances by minimizing the distance of prototypes with the same category but from different domains. Moreover, to avoid the non-convergence problem during multi-level feature alignment, a domain-consistent loss is proposed to harmonize the adaptation training process. Comprehensive results on various cross-domain detection tasks demonstrate the broad applicability and effectiveness of the proposed approach.

关键词： cross-domain object detection hierarchical feature alignment domain-consistent loss foreground-aware attention adversarial learning

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：