检索结果-内蒙古大学图书馆

FPDIoU Loss: A loss function for efficient bounding box regression of rotated object detection

IMAGE AND VISION COMPUTING 2025年 154卷

作者： Ma, Siliang Xu, Yong South China Univ Technol Inst Comp Sci & Engn 382Outer Ring East Rd Guangzhou 510006 Guangdong Peoples R China Pengcheng Lab 2Xingke 1st St Shenzhen 518000 Guangdong Peoples R China

Bounding box regression is one of the important steps of object detection. However, rotation detectors often involve amore complicated loss based on SkewIoU which is unfriendly to gradient-based training. Most of the existing loss functions for rotated object detection calculate the difference between two bounding boxes only focus on the deviation of area or each points distance (e.g., G h - 1 , G and G ). The calculation process of some loss functions is extremely complex (e.g. G ). In order to improve the efficiency and accuracy of bounding box regression for rotated object detection, we proposed a novel metric for arbitrary shapes comparison based on minimum points distance, which takes most of the factors from existing loss functions for rotated object detection into account, i.e., the overlap or nonoverlapping area, the central points distance and the rotation angle. We also proposed a loss function called G based on four points distance for accurate bounding box regression focusing on faster and high quality anchor boxes. In the experiments, loss has been applied to state-of-the-art rotated object detection (e.g., RTMDET, H2RBox) models training with three popular benchmarks of rotated object detection including DOTA, DIOR, HRSC2016 and two benchmarks of arbitrary orientation scene text detection including ICDAR 2017 RRC-MLT and ICDAR 2019 RRC-MLT, which achieves better performance than existing loss functions. The code is available at https://***/JacksonMa618/FPDIoU

关键词： rotated object detection Bounding box regression Loss function Minimum points distance

来源：评论

学校读者我要写书评

暂无评论

Explicit and Implicit Box Equivariance Learning for Weakly-Supervised rotated object detection

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2025年第1期9卷 509-521页

作者： Wang, Linfei Zhan, Yibing Lin, Xu Yu, Baosheng Ding, Liang Zhu, Jianqing Tao, Dapeng Yunnan Univ Kunming 650091 Yunnan Peoples R China Yunnan Key Lab Media Convergence Kunming 650091 Yunnan Peoples R China JD Explore Acad Beijing 100000 Peoples R China Yunnan Union Vis Innovat Technol Co Ltd Kunming 650500 Yunnan Peoples R China Univ Sydney Camperdown NSW 2006 Australia Huaqiao Univ Coll Engn Quanzhou 362021 Peoples R China

rotated object detection has emerged as a preferred alternative to the conventional Horizontal Box (HBox) object detection in various tasks. However, the majority of existing detection datasets are still annotated using HBox. Re-annotating these datasets for rotated object detection is notably inefficient and labor-intensive. Recently, weakly-supervised detectors based directly on HBox have provided a solution to bridge the gap between the rotated bounding box (RBox) prediction and the weak annotations. Nonetheless, they tend to overlook the potential impact of factors such as bounding box position, aspect ratio, and scale on angle prediction. Additionally, inconsistencies arise when evaluating rotation variations, hindering the inference performance. In light of these limitations, we propose an explicit and implicit equivariance detector (EIE-Det) to predict fine-grained RBox based on HBox annotations. Along with a shared backbone and neck structure, our proposed EIE-Det comprises a weakly supervised (WS) branch for predicting coarse RBox from HBox labels, an explicit equivariance (EE) branch for learning rotation consistency, and an implicit equivariance (IE) branch for learning position, aspect ratio, and scale consistency. Furthermore, we introduce the Tanimoto Coefficient for vector-based bounding box regression, enabling a comprehensive assessment of model prediction bias, thereby eliminating interference from variations in box aspect ratio and scale. Experiments demonstrate that our method not only significantly surpasses state-of-the-art models supervised with HBox but also achieves accuracy comparable to fully-supervised detectors (75.74% and 58.50% AP(50) on DOTA and DIOR-R, respectively).

关键词： object detection Feature extraction Detectors Annotations Task analysis Neck Shape Box equivariance rotated object detection weakly-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing rotated object detection via anisotropic Gaussian bounding box and Bhattacharyya distance

引用

NEUROCOMPUTING 2025年 623卷

作者： Thai, Chien Trang, Mai Xuan Ninh, Huong Ly, Hoang Hiep Le, Anh Son Phenikaa Univ Sch Comp Hanoi 12116 Vietnam Phenikaa X Joint Stock Co Phenikaa Grp Hanoi Vietnam Viettel Grp Viettel Aerosp Inst Optoelect Ctr Hanoi Vietnam Hanoi Univ Sci & Technol Sch Mech Engn Dept Mechatron Hanoi Vietnam Phenikaa Univ Fac Vehicle & Energy Engn Hanoi 12116 Vietnam

Detecting rotated objects accurately and efficiently is a significant challenge in computer vision, particularly in applications such as aerial imagery, remote sensing, and autonomous driving. Although traditional object detection frameworks are effective for axis-aligned objects, they often underperform in scenarios involving rotated objects due to their limitations in capturing orientation variations. This paper introduces an improved loss function aimed at enhancing detection accuracy and robustness by leveraging the Gaussian bounding box representation and Bhattacharyya distance. In addition, we advocate for the use of an anisotropic Gaussian representation to address the issues associated with isotropic variance in square-like objects. Our proposed method addresses these challenges by incorporating a rotation-invariant loss function that effectively captures the geometric properties of rotated objects. We integrate this proposed loss function into state-of-the-art deep learning-based rotated object detection detectors, and extensive experiments demonstrated significant improvements in mean Average Precision metrics compared to existing methods. The results highlight the potential of our approach to establish anew benchmark in rotated object detection, with implications fora wide range of applications that require precise and reliable object localization irrespective of orientation.

关键词： rotated object detection Bounding box regression Gaussian distribution Bhattacharyya distance

来源：评论

学校读者我要写书评

暂无评论

Gaussian-based R-CNN with large selective kernel for rotated object detection in remote sensing images

引用

NEUROCOMPUTING 2025年 620卷

作者： Yang, Xiao Mohamed, Ahmad Sufril Azlan Univ Sains Malaysia Sch Comp Sci Minden 11800 Penang Malaysia Zhejiang Polytech Univ Mech & Elect Engn Hangzhou 310059 Peoples R China

Accurate object detection in remote sensing images is essential for applications such as environmental monitoring, urban planning, and disaster management. However, the detection of small rotated objects remains a major challenge, limiting advancements in remote sensing object detection. The performance of generic object detectors tends to drastically deteriorate on small rotated object detection tasks. Generic object detectors often struggle with tasks involving small rotated objects due to issues such as feature misalignment, angle periodicity, and metric-loss inconsistency, arising from arbitrary orientations and low-quality feature representations caused by their small sizes. In response to these obstacles, we propose a Gaussian-based R-CNN, providing a unified solution to the complexities of rotated object detection. Our approach integrates Oriented Region Proposal Network and rotated Region of Interest Align to extract rotation-invariant features. Additionally, Gaussian Modeling is introduced to transform oriented bounding boxes into Gaussian distributions, effectively avoiding the regression of the angle parameters. Moreover, we introduce a Dynamic Center-Aligned Gaussian Product Loss (DCAGPL) to accurately measure the similarity between Gaussian distributions, approximating rotated Intersection over Union loss. Furthermore, our method incorporates large selective kernel attention, dynamically adjusting its large spatial receptive field to enhance the modeling of the surrounding context for various objects, particularly crucial for the detection of small objects. Extensive experiments validate the effectiveness of our proposed methods. Without any additional tricks, the Gaussian-based R-CNN with ResNet50 achieves state-of-the-art detection accuracy on two commonly used datasets for rotated object detection: DOTA (75.99% mAP) and HRSC2016-MS (68.02% mAP). Furthermore, when integrated with the large selective kernel mechanism, it achieves mAP scores of 76.77% a

关键词： rotated object detection Gaussian Modeling Large kernel Selective attention Remote sensing

来源：评论

学校读者我要写书评

暂无评论

PARDet: Dynamic point set alignment for rotated object detection

引用

PATTERN RECOGNITION 2024年 153卷

作者： Xu, Yihao Shen, Jifeng Dai, Ming Yang, Wankou Southeast Univ Sch Automat Nanjing 210096 Peoples R China Jiangsu Univ Sch Elect & Informat Engn Zhenjiang 212013 Peoples R China

Due to the inherent mismatch between rotated objects and horizontal features, feature point misalignment has been a challenge in the task of rotated object detection (ROD). Specifically, considering the pattern of convolution, foreground features are often mixed with background noise, which can confuse the model and affect the model from feature point alignment during the training phase. To mitigate this issue, previous methods concentrate on fixed positions derived from predicted boxes by additionally introducing a refinement stage. However, merely learning fixed position priors during training can result in suboptimal alignment and inefficiency during inference. This paper introduces a dynamic point alignment detector to concurrently address issues associated with feature misalignment and inference inefficiency. The method is made up of two components: the fine-grained points generator (FPG) captures key information, and the point alignment module (PAM) derives precise feature representations. Both modules empower the detector with the capability to dynamically perceive rotated objects, extracting more comprehensive and reasoned feature contents from the feature maps. In general, our method enables the model to independently identify and prioritize the valuable features during the training process. Subsequently, during the inference stage, the results can be directly predicted without additional alignment operations. The experimental results demonstrate that our method can achieve competitive and superior results with average precision (AP) values of 79.33%, 95.73%, and 63.37% on the DOTA, HRSC2016, and DIOR-R datasets, respectively. Codes will be publicly available at https://***/Xuyihaoby/PARDet.

关键词： Remote sensing rotated object detection Fine-grained feature Feature alignment

来源：评论

学校读者我要写书评

暂无评论

Learning power Gaussian modeling loss for dense rotated object detection in remote sensing images

引用

Chinese Journal of Aeronautics 2023年第10期36卷 353-365页

作者： Yang LI Haining WANG Yuqiang FANG Shengjin WANG Zhi LI Bitao JIANG Department of Space Information Space Engineering UniversityBeijing 101416China Beijing Institute of Remote Sensing Information Beijing 100192China Department of Electronic Engineering Tsinghua UniversityBeijing 100084China

object detection in Remote Sensing(RS)has achieved tremendous advances in recent years,but it remains challenging for rotated object detection due to cluttered backgrounds,dense object arrangements and the wide range of size variations among *** tackle this problem,Dense Context Feature Pyramid Network(DCFPN)and a powerα-Gaussian loss are designed for rotated object detection in this *** proposed DCFPN can extract multi-scale information densely and accurately by leveraging a dense multi-path dilation layer to cover all sizes of objects in remote sensing *** more accurate detection while avoiding bottlenecks such as boundary discontinuity in rotated bounding box regression,a-Gaussian loss,a unified power generalization of existing Gaussian modeling losses is ***,the properties ofα-Gaussian loss are analyzed comprehensively for a wider range of *** results on four datasets(UCAS-AOD,HRSC2016,DIOR-R,and DOTA)show the effectiveness of the proposed method using different detectors,and are superior to the existing methods in both feature extraction and bounding box regression。

关键词： Convolutional neural networks Distribution metric Losses Remote sensing rotated object detection

来源：评论

学校读者我要写书评

暂无评论

rotatedStomataNet: a deep rotated object detection network for directional stomata phenotype analysis

引用

PLANT CELL REPORTS 2024年第5期43卷 126-126页

作者： Yang, Xiaohui Wang, Jiahui Li, Fan Zhou, Chenglong Wu, Minghui Zheng, Chen Yang, Lijun Li, Zhi Li, Yong Guo, Siyi Song, Chunpeng Henan Univ Henan Engn Res Ctr Artificial Intelligence Theory Sch Math & Stat Kaifeng 475004 Henan Peoples R China Cent South Univ Sch Automat Changsha 410000 Hunan Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Wuxi 214400 Jiangsu Peoples R China Henan Univ State Key Lab Crop Stress Adaptat & Improvement Kaifeng 475004 Henan Peoples R China Henan Univ Acad Adv Interdisciplinary Studies Kaifeng 475004 Henan Peoples R China

Key MessageInnovatively, we consider stomatal detection as rotated object detection and provide an end-to-end, batch, rotated, real-time stomatal density and aperture size intelligent detection and identification system, *** acts as a pathway for air and water vapor in the course of respiration, transpiration, and other gas metabolism, so the stomata phenotype is important for plant growth and development. Intelligent detection of high-throughput stoma is a key issue. Nevertheless, currently available methods usually suffer from detection errors or cumbersome operations when facing densely and unevenly arranged stomata. The proposed rotatedStomataNet innovatively regards stomata detection as rotated object detection, enabling an end-to-end, real-time, and intelligent phenotype analysis of stomata and apertures. The system is constructed based on the Arabidopsis and maize stomatal data sets acquired destructively, and the maize stomatal data set acquired in a non-destructive way, enabling the one-stop automatic collection of phenotypic, such as the location, density, length, and width of stomata and apertures without step-by-step operations. The accuracy of this system to acquire stomata and apertures has been well demonstrated in monocotyledon and dicotyledon, such as Arabidopsis, soybean, wheat, and maize. The experimental results that the prediction results of the method are consistent with those of manual labeling. The test sets, the system code, and their usage are also given (https://***/AITAhenu/rotatedStomataNet).

关键词： Phenotypic analysis Stoma Aperture Computer vision rotated object detection

来源：评论

学校读者我要写书评

暂无评论

Orientation-First Strategy With Angle Attention Module for rotated object detection in Remote Sensing Images

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2022年 15卷 8492-8505页

作者： Zhang, Yuxi Wang, Yongcheng Zhang, Ning Li, Zheng Zhao, Zhikang Gao, Yunxiao Xu, Dongdong Ben, Guangli Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys Changchun 130033 Peoples R China Univ Chinese Acad Sci Beijing 100049 Peoples R China

Recently, object detection in remote sensing images (RSIs) have received extensive attention and made significant progress. Nonetheless, the arbitrary orientations of objects in RSIs make their detection a challenging task. Most of the existing detection methods are difficult to extract the orientation features of objects due to the lack of directionality of conventional convolutions. In addition, the boundary discontinuity in angle regression affects the detection of object orientations. In response to these problems, this article proposes an orientation-first refinement detector (OFRDet), which is based on a strategy that enables the detector to detect the angle of an object ahead of others and presets oriented anchors. In OFRDet, we propose an angle encoding regression module (AERM) and an angle channel attention module (ACAM). AERM transforms angle detection into multiparameter regression, which eliminates boundary discontinuities. ACAM uses convolution kernels with different angles to extract directional features purposefully according to the preset oriented anchors. After these two modules, more accurate bounding boxes are generated and sent to the refined stage to obtain the final detection results. We evaluate our method and demonstrate the effectiveness of it by conducting experiments on two challenging and credible datasets, DOTA, HRSC2016. OFRDet achieves competitive results 79.56%, 96.29% mAP on the two datasets, respectively.

关键词： Detectors object detection Feature extraction Remote sensing Redundancy Kernel Encoding Angle channel attention angle encoding remote sensing images rotated object detection

来源：评论

学校读者我要写书评

暂无评论

RSDet plus plus : Point-Based Modulated Loss for More Accurate rotated object detection

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2022年第11期32卷 7869-7879页

作者： Qian, Wen Yang, Xue Peng, Silong Zhang, Xiujuan Yan, Junchi Chinese Acad Sci Inst Automat Beijing 100190 Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing 101408 Peoples R China Shanghai Jiao Tong Univ Dept Comp Sci & Engn Shanghai 200240 Peoples R China Inner Mongolia Key Lab Mol Biol Featured Plants Hohhot 010018 Peoples R China

We classify the discontinuity of loss in both five-param and eight-param rotated object detection methods as rotation sensitivity error (RSE) which will result in performance degeneration. We introduce a novel modulated rotation loss to alleviate the problem and a rotation sensitivity detection network (RSDet) which consists of an eight-param single-stage rotated object detector and the modulated rotation loss. Our proposed RSDet has several advantages: 1) it reformulates the rotated object detection problem as predicting the corners of objects while most previous methods employ a five-param-based regression method with different measurement units. 2) modulated rotation loss achieves consistent improvement on both five-param and eight-param rotated object detection methods by solving the discontinuity of loss. To further improve the accuracy of our method on objects smaller than 10 pixels, we introduce a novel RSDet++ which consists of a point-based anchor-free rotated object detector and a modulated rotation loss. Extensive experiments demonstrate the effectiveness of both RSDet and RSDet++, which achieve competitive results on rotated object detection in the challenging benchmarks DOTA-v1.0, DOTA-v1.5, and DOTA-v2.0. We hope the proposed method can provide a new perspective for designing algorithms to solve rotated object detection and pay more attention to tiny objects. The codes and models are available at: https://***/yangxue0827/Rotationdetection.

关键词： object detection Detectors Sensitivity Feature extraction Benchmark testing Training Measurement units rotated object detection modulated loss point-based tiny objects

来源：评论

学校读者我要写书评

暂无评论

Acceleration of rotated object detection on FPGA

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS 2022年第4期69卷 2296-2300页

作者： Fan, Xitian Xie, Guangwei Huang, Zhongchen Cao, Wei Wang, Lingli Fudan Univ Sch Comp Sci Shanghai 200433 Peoples R China Shanghai HONGZHEN Informat Sci & Technol Corp Dept Comp Res Shanghai Peoples R China Fudan Univ Inst Big Data Shanghai 200433 Peoples R China Fudan Univ State Key Lab ASIC & Syst Shanghai 200433 Peoples R China

In this brief, an FPGA-based solution is proposed to show the computing efficiency on rotated object detection based on R(3)Det algorithm. The key idea of our approach is firstly to design reconfigurable neural processing units (NPU) for convolutional neural networks (CNN) and a specific architecture for spatial operations, and then to adopt a novel scheduling scheme to deal with the data dependency on these modules. When implemented on cost-effective Kintex Ultrascale+ XCKU15P FPGA, the proposed solution achieves nearly 3.20 times energy efficiency compared to NVIDIA V100 GPU. The proposed reconfigurable NPU achieves 73.20% DSP efficiency, 30.39% improvement compared to the 56.14% DSP efficiency of the state-of-the-art CNN accelerator on ResNet-50.

关键词： Graphics processing units Computer architecture object detection Field programmable gate arrays Neural networks Convolutional neural networks Systems architecture rotated object detection acceleration architecture CNN FPGA

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：