Learning Rotation-Invariant and Fisher Discriminative Convolutional Neural Networks for Object Detection
The DOI matches its Crossref record; 71 of the 77 references it cites were found in a registry or an index.
Abstract
The performance of object detection has recently been significantly improved due to the powerful features learnt through convolutional neural networks (CNNs). Despite the remarkable success, there are still several major challenges in object detection, including object rotation, within-class diversity, and between-class similarity, which generally degenerate object detection performance. To address these issues, we build up the existing state-of-the-art object detection systems and propose a simple but effective method to train rotation-invariant and Fisher discriminative CNN models to further boost object detection performance. This is achieved by optimizing a new objective function that explicitly imposes a rotation-invariant regularizer and a Fisher discrimination regularizer on the CNN features. Specifically, the first regularizer enforces the CNN feature representations of the training samples before and after rotation to be mapped closely to each other in order to achieve rotation-invariance. The second regularizer constrains the CNN features to have small within-class scatter but large between-class separation. We implement our proposed method under four popular object detection frameworks, including region-CNN (R-CNN), Fast R- CNN, Faster R- CNN, and R- FCN. In the experiments, we comprehensively evaluate the proposed method on the PASCAL VOC 2007 and 2012 data sets and a publicly available aerial image data set. Our proposed methods outperform the existing baseline methods and achieve the state-of-the-art results.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Designing rotation invariant loss functions is a recent popular solution, where a constraint on the detection loss is added so that the feature of rotated objects keeps unchanged [169-171].”
From Object Detection in 20 Years: A Survey (2023) · cites this paper as [171]
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 8×
“In Table 10, the baseline method is the combination of R-FCN [5] and ResNet-101 [9].”
InfluentialVery Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 9×
“In Table 7, the baseline method is the combination of the R-CNN framework [1], [2], VGGNet-16 [13], and bounding box regression (BB).”
InfluentialImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 6×
“We implement our method for aerial car detection within the R-CNN framework and adopt the most popular AlexNet CNN [6], which is pre-trained on ImageNet [7], as our basic network structure.”
InfluentialFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 10×
“To further reduce the computational costs of R-CNN [1], [2] and Fast R-CNN [3], Faster R-CNN [4] was proposed.”
InfluentialFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812 · cited 10×
“To further reduce the computational costs of R-CNN [1], [2] and Fast R-CNN [3], Faster R-CNN [4] was proposed.”
InfluentialImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 7×
“Then, R-CNN extracts CNN features from each object proposal and classifies them as objects or non-objects with class-specific support vector machines (SVMs), where the deep CNN model was pre-trained on ImageNet data set [7] and then fine-tuned on a small set of images annotated for the detection task.”
InfluentialYou Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380 · cited 1×
“More recently, due to the great success of deep convolutional neural networks (CNNs) for image classification [6] on the ImageNet large scale visual recognition challenge (ILSVRC) [7], [8], CNN-based methods for object detection have been attracting increasing research interest because these methods have achieved stat…”
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
Research works printed but not linked: 3
- lin, 2013, Network in Network
- he, 2014, Spatial pyramid pooling in deep convolutional networks for visual recognition, Proc Eur Conf Comput Vis, 346
- girshick, 2015, Fast R-CNN, Proc IEEE Int Conf Comput Vis, 1440
Cited by
Items 101–200 of 408
Vehicle Detection From High-Resolution Remote Sensing Imagery Using Convolutional Capsule Networks
Yu, Y., Gu, T., Guan, H., Li, D., & Jin, S. (2019) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2019.2912582 · Cited by 58
“In [22], rotation-invariant and Fisher discriminative CNNs were trained to handle rotations, within-class diversity, and between-class similarity.”
MethodA deep learning approach for face recognition based on angularly discriminative features
Iqbal, M., Sameem, M. S. I., Naqvi, N., Kanwal, S., & Ye, Z. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2019.10.002 · Cited by 59
Vision-to-Language Tasks Based on Attributes and Attention Mechanism
Li, X., Yuan, A., & Lu, X. (2019) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2019.2914351 · Cited by 62
“With the development of deep learning, some related and recent work on deep learning has been researched for visual content analysis [21]–[23].”
BackgroundSmooth GIoU Loss for Oriented Object Detection in Remote Sensing Images
Qian, X., Zhang, N., & Wang, W. (2023) · Remote Sensing · doi:10.3390/rs15051259 · Cited by 53
Li, W., Hsu, C.-Y., & Hu, M. (2021) · Annals of the American Association of Geographers · doi:10.1080/24694452.2021.1877527 · Cited by 52
“In addition, weakly supervised learning and target detection from remote sensing image has been exploited by geospatial researchers.”
Hao, S., Wu, B., Zhao, K., Ye, Y., & Wang, W. (2022) · Remote Sensing · doi:10.3390/rs14061507 · Cited by 55
Monitoring and Identification of Road Construction Safety Factors via UAV
Zhu, C., Zhu, J., Bu, T., & Gao, X. (2022) · Sensors · doi:10.3390/s22228797 · Cited by 61
“Convolutional neural networks [30,31] are one of the most widely used and effective tools.”
MethodMa, A., Yu, N., Zheng, Z., Zhong, Y., & Zhang, L. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3151405 · Cited by 60
“It involves labeling the given scene images with predefined semantic categories [1], [2], and has been widely used in urban planning [3], [4], geological disaster monitoring [5], [6], remote sensing image retrieval [7]–[9], and other fields [10], [11].”
Learning Discriminative Representation for Fine-Grained Object Detection in Remote Sensing Images
Xie, X., Cheng, G., Li, W., Lang, C., Zhang, P., Yao, Y., & Han, J. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3544741 · Cited by 60
“…and OA are Boeing 737, Boeing 777, Boeing 747, Boeing 787, Airbus A321, Airbus A220, Airbus A330, Airbus A350, COMAC C919, COMAC ARJ21, and other airplanes. has become an urgent research topic in the community [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19].”
Self-citationLearning From Synthetic Images via Active Pseudo-Labeling
Song, L., Xu, Y., Zhang, L., Du, B., Zhang, Q., & Wang, X. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.2989100 · Cited by 56
“…learning techniques [1] and their successful applications for many challenging computer vision tasks, including image classification [2]–[4], object detection [5], [6], semantic segmentation [7]–[9], super resolution [10], [11], image quality enhancement [12]–[15], re-identification [16], [17], etc .”
BackgroundAFAN: Augmented Feature Alignment Network for Cross-Domain Object Detection
Wang, H., Liao, S., & Shao, L. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3066046 · Cited by 61
“Other attempts of object detection include learning rotation-invariant features [25], self-supervised feature augmentation [26].”
Object Detection for Aerial Images With Feature Enhancement and Soft Label Assignment
Yu, Y., Yang, X., Li, J., & Gao, X. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3177255 · Cited by 60
“Object detection in aerial images has drawn great attention recently which aims to localize the objects in images and identify their categories simultaneously [1]–[5].”
BackgroundBahri, A., Majelan, S. G., Mohammadi, S., Noori, M., & Mohammadi, K. (2019) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2019.2937872 · Cited by 57
Carranza-García, M., Lara-Benítez, P., García–Gutiérrez, J., & Riquelme, J. C. (2021) · Neurocomputing · doi:10.1016/j.neucom.2021.04.001 · Cited by 53
“[22] Due to the cost of manually labeling bounding boxes, there are also many studies on weakly-supervised object detection that work only with image-level labels [23].”
BackgroundMultichannel Pulse-Coupled Neural Network-Based Hyperspectral Image Visualization
Duan, P., Kang, X., Shutao Li, A., & Ghamisi, P. (2019) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2019.2949427 · Cited by 48
“Digital Object Identifier 10.1109/TGRS.2019.2949427 extensively applied in various aspects [1]–[9], including object recognition [10], [11], land-cover classification [12], [13], and anomaly detection [14], [15].”
BackgroundInfluentialTraining Robust Object Detectors From Noisy Category Labels and Imprecise Bounding Boxes
Xu, Y., Zhu, L., Yang, Y., & Wu, F. Q. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3085208 · Cited by 55
“Most previous methods [31]–[37] treat weakly supervised object detection as a multi-instance learning problem.”
MethodYuan, Z., Liu, Z., Zhu, C., Qi, J., & Zhao, D. (2021) · Remote Sensing · doi:10.3390/rs13050862 · Cited by 50
“Many optical remote sensing object detection methods today use R-CNN based networks [27,29].”
BackgroundMethodInfluentialIPGN: Interactiveness Proposal Graph Network for Human-Object Interaction Detection
Wang, H., Jiao, L., Liu, F., Li, L., Liu, X., Ji, D., & Gan, W. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3096333 · Cited by 46
Liu, P., Wang, Q., Zhang, H., Mi, J., & Liu, Y. (2023) · Remote Sensing · doi:10.3390/rs15092429 · Cited by 52
An Improved InceptionV3 Network for Obscured Ship Classification in Remote Sensing Images
Liu, K., Yu, S., & Liu, S. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.3017676 · Cited by 45
Local Attention Networks for Occluded Airplane Detection in Remote Sensing Images
Zhou, M., Zou, Z., Shi, Z., Zeng, W., & Gui, J. (2019) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2019.2924822 · Cited by 47
Hou, B., Ren, Z., Zhao, W., Wu, Q., & Jiao, L. (2019) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2019.2942103 · Cited by 48
“Rotation-invariant and Fisher discriminative CNN models are put forward in [20] to alleviate the performance degradation caused by object rotation, within-class diversity, and between-class similarity.”
BackgroundA Lightweight Object Detection Framework for Remote Sensing Images
Huyan, L., Bai, Y., Li, Y., Jiang, D., Zhang, Y., Zhou, Q., Wei, J., Liu, J., Zhang, Y., & Cui, T. (2021) · Remote Sensing · doi:10.3390/rs13040683 · Cited by 48
“Besides, rotation-invariant regularity and Fisher discrimination regularizer [65] are added to Faster-RCNN.”
MethodÖzbay, E., & Ozbay, F. A. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.106574 · Cited by 46
Generalization-Enhanced Few-Shot Object Detection in Remote Sensing
Lin, H., Li, N., Yao, P., Dong, K., Guo, Y., Hong, D., Zhang, Y., & Wen, C. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3528262 · Cited by 51
“On the other hand, two-stage methods [6], [16], [17], [18] first generate region proposals in the initial stage and then classify these proposals in the second stage.”
Dependency-Aware Attention Control for Image Set-Based Face Recognition
Liu, X., Guo, Z., You, J., & Kumar, B. V. K. V. (2019) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2019.2938418 · Cited by 44
A Lightweight Adaptive RoI Extraction Network for Precise Aerial Image Instance Segmentation
Zeng, X., Wei, S., Shi, J., & Zhang, X. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3121485 · Cited by 45
Xu, C., Zhu, G., & Shu, J. (2022) · Remote Sensing · doi:10.3390/rs14061445 · Cited by 46
Su, Z., Yu, S., Chu, J., Zhai, Q., Gong, J., & Fan, H. (2020) · Advanced Engineering Informatics · doi:10.1016/j.aei.2020.101055 · Cited by 44
Object Detection for Unmanned Aerial Vehicle Camera via Convolutional Neural Networks
Saetchnikov, I. V., Tcherniavskaia, E. A., & Skakun, V. V. (2020) · IEEE Journal on Miniaturization for Air and Space Systems · doi:10.1109/jmass.2020.3040976 · Cited by 49
“…75.92% accuracy [19], etc. Adaptation of the region convolution neural network (R-CNN), Fast R-CNN, Faster R-CNN, and R-FCN architectures for object detection tasks has been recently addressed via the rotation-invariant and Fisher discriminative convolutional neural network (CNN) models [20].”
MethodInfluentialSelf-Guided Proposal Generation for Weakly Supervised Object Detection
Cheng, G., Xie, X., Chen, W., Feng, X., Yao, X., & Han, J. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3181466 · Cited by 45
“[8] proposed a multi-scale object proposal network (MS-OPN) and an accurate object detection network (AODN) for simultaneously detecting multi-class objects in remote sensing images with large scales variability.”
BackgroundSelf-citationMSSDet: Multi-Scale Ship-Detection Framework in Optical Remote-Sensing Images and New Benchmark
Chen, W., Han, B., Zheng, Y., & Gao, X. (2022) · Remote Sensing · doi:10.3390/rs14215460 · Cited by 56
MM-RCNN: Toward Few-Shot Object Detection in Remote Sensing Images With Meta Memory
Li, J., Tian, Y. W., Xu, Y., Hu, X., Zhang, Z., Hu, W., & Xiao, Y. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3228612 · Cited by 49
“For example, Cheng et al. [18] presented RIFD-CNN to explicitly address the problems related to object orientation, intraclass diversity, and interclass similarity.”
BackgroundMethodSSSDET: Simple Short and Shallow Network for Resource Efficient Vehicle Detection in Aerial Scenes
Mandal, M., Shah, M., Meena, P., & Vipparthi, S. K. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803262 · Cited by 43
“In addition, to effectively deal with the challenges of rotation variations and appearance ambiguity in geospatial scenes, various rotation-invariant detectors [3, 16-18] have been proposed in the literature.”
BackgroundRobust Vehicle Detection in High-Resolution Aerial Images With Imbalanced Data
Li, X., Li, X., Li, Z., Xiong, X., Khyam, M. O., & Sun, C. (2021) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2021.3081057 · Cited by 43
Feng, X., Yao, X., Cheng, G., Han, J., & Han, J. (2021) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2021.3105575 · Cited by 42
Self-citationCNN-Based Copy-Move Forgery Detection Using Rotation-Invariant Wavelet Feature
Lee, S. I., Park, J. Y., & Eom, I. K. (2022) · IEEE Access · doi:10.1109/access.2022.3212069 · Cited by 40
“To address this problem, various attempts have been made to create CNN rotation-invariance [41], [42].”
MethodDegradation Modeling for Restoration-Enhanced Object Detection in Adverse Weather Scenes
Wang, X., Liu, X., Yang, H. W., Wang, Z., Wen, X., He, X., Qing, L., & Chen, H. (2024) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2024.3442924 · Cited by 53
“Existing object detection algorithms [37], [38], [39], [40], [41], [42] can generally be divided into two categories based on the type of network: one is based on region proposal methods, and the other is based on regression methods.”
Rotated Feature Network for Multiorientation Object Detection of Remote-Sensing Images
Zhou, K., Zhang, Z., Gao, C., & Liu, J. (2020) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2020.2965629 · Cited by 38
“Cheng et al. [5] presented an effective method to learn rotation-invariant features, successfully applied where object orientations vary a lot.”
MethodJiang, S., Yao, W., Wong, M. S., Li, G., Hong, Z., Kuc, T., & Tong, X. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.2975606 · Cited by 40
“To adapt the aerial view, an extended version of VGG [26] were reported trained by using extra aerial view datasets, such as NWPU-RESISC45 [27].”
MethodGuiding Clean Features for Object Detection in Remote Sensing Images
Cheng, G., He, M., Hong, H., Yao, X., Qian, X., & Guo, L. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3104112 · Cited by 46
Self-citation3D IoU-Net: IoU Guided 3D Object Detector for Point CloudsPreprint
Li, J., Luo, S., Zhu, Z., Dai, H., Krylov, A. S., Ding, Y. S., & Shao, L. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.04962 · Cited by 62
MashFormer: A Novel Multiscale Aware Hybrid Detector for Remote Sensing Object Detection
Wang, K., Bai, F., Li, J., Liu, Y., & Li, Y. (2023) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2023.3254047 · Cited by 40
“[23] added a rotation-invariant layer and a Fisher discriminative layer to an existing CNN model to deal with object rotation.”
BackgroundAdegun, A. A., Viriri, S., & Tapamo, J. (2024) · Applied Intelligence · doi:10.1007/s10489-024-05818-y · Cited by 38
Using multi-label classification to improve object detection
Gong, T., Liu, B., Chu, Q., & Yu, N. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.08.089 · Cited by 37
Fang, G., Ba, S., Gu, Y., Lin, Z., Hou, Y., Qin, C., Zhou, C.-C., Xu, J., Dai, Y., Song, J., et al. (2023) · The Astronomical Journal · doi:10.3847/1538-3881/aca1a6 · Cited by 39
Adaptive Saliency Biased Loss for Object Detection in Aerial Images
Sun, P., Chen, G., & Shang, Y. (2020) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.2980023 · Cited by 34
“For example, Cheng et al. [30], [31] proposed rotation-invariant deep learning models with regularization to achieve state-of-the-art performance on remote sensing images.”
BackgroundGarajeh, M. K., Blaschke, T., Haghi, V. H., Weng, Q., Kamran, K. V., & Li, Z. (2022) · Canadian Journal of Remote Sensing · doi:10.1080/07038992.2022.2056435 · Cited by 34
“Several algorithms have been proposed to optimize the classification results in DL-CNN, including the SGD (Stochastic Gradient Descent), the SGD þ Momentum, the Adam (Adaptive Moment Optimization), the Adagrad, and the Adadelta algorithms (Gong et al. 2019).”
MethodKurdthongmee, W. (2020) · Heliyon · doi:10.1016/j.heliyon.2020.e03480 · Cited by 33
TOSO: Student’s-T Distribution Aided One-Stage Orientation Target Detection in Remote Sensing Images
Feng, P., Lin, Y., Guan, J., He, G., Shi, H., & Chambers, J. A. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053562 · Cited by 35
“Recently, extensive studies have been devoted to object detection in remote sensing images [1, 2, 3, 4, 5, 6, 7, 8].”
BackgroundFast Multi-Focus Ultrasound Image Recovery Using Generative Adversarial Networks
Goudarzi, S., Asif, A. A., & Rivaz, H. J. (2020) · IEEE Transactions on Computational Imaging · doi:10.1109/tci.2020.3019137 · Cited by 34
“Defining a proper objective function to minimize during the training phase is a critical factor that influences the performance of the network, and is currently an active field of research [11], [12], [13].”
BackgroundGajbhiye, G. O., & Nandedkar, A. V. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105076 · Cited by 35
A Coarse-to-Fine Network for Ship Detection in Optical Remote Sensing Images
Wu, Y., Ma, W., Gong, M., Bai, Z., Zhao, W., Guo, Q., Chen, X., & Miao, Q. (2020) · Remote Sensing · doi:10.3390/rs12020246 · Cited by 30
“[40] trained a classification network to detect the locations of ship heads, and adopted an iterative multitask network to perform bounding-box regression and classification [41].”
MethodSeverity-Aware Semantic Segmentation With Reinforced Wasserstein Training
Liu, X., Ji, W., You, J., Fakhri, G. E., & Woo, J. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.01258 · Cited by 31
“Our claim is also closely related to the importance-aware segmentation/classification [7, 10, 26].”
BackgroundYao, J., Liang, W., & Xiong, J. (2022) · International Journal of Pressure Vessels and Piping · doi:10.1016/j.ijpvp.2022.104781 · Cited by 31
Vehicle Detection in Overhead Satellite Images Using a One-Stage Object Detection Model
Stuparu, D.-G., Ciobanu, R., & Dobre, C. (2020) · Sensors · doi:10.3390/s20226485 · Cited by 34
“Other novel and interesting neural network-based solutions for object detection in aerial images are proposed and presented in [12,13], but the purpose is somewhat more general than the goal of our paper, since the authors focus on multiple classes of aerial images and not particularly on vehicles.”
BackgroundRotated cascade R-CNN: A shape robust detector with coordinate regression
Zhu, Y., Ma, C., & Du, J. (2019) · Pattern Recognition · doi:10.1016/j.patcog.2019.106964 · Cited by 31
ℱ3-Net: Feature Fusion and Filtration Network for Object Detection in Optical Remote Sensing Images
Ye, X., Xiong, F., Lu, J., Zhou, J. W., & Qian, Y. (2020) · Remote Sensing · doi:10.3390/rs12244027 · Cited by 33
“To this end, many researches attempt to tackle object detection in RS images by considering the above unique characteristics of RS objects and embedding related factors into the existing networks [16–18].”
BackgroundMTFFNet: a Multi-task Feature Fusion Framework for Chinese Painting Classification
Jiang, W., Wang, X., Ren, J., Li, S., Sun, M., Wang, Z., & Jin, J. S. (2021) · Cognitive Computation · doi:10.1007/s12559-021-09896-9 · Cited by 34
“Recently, the successful application of deep learning technology in the tasks of handwritten digit recognition [12], human motion recognition [13], speech recognition [14], saliency detection [15], object detection [16], and ImageNet [17]-based classification has shown its considerably powerful ability in extracting d…”
BackgroundOptical remote sensing image object detection based on multi-resolution feature fusion
Yao, Y., Gong, C., Xie, X., & Junwei, H. (2021) · National Remote Sensing Bulletin · doi:10.11834/jrs.20210505 · Cited by 29
Qian, X., Chen, X., Yue, W., Liu, X., Guo, J., Li, Z., Li, Y., & Wang, W. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.3019582 · Cited by 31
“The high-resolution remote sensing image scene classi-fication is to distinguish the land use or coverage category of remote sensing image according to the image content [3], [4], which can provide the important cues for other remote sensing image processing tasks [5]–[10].”
BackgroundOu, Y., Tsai, A., Zhou, X., & Wang, J.-F. (2019) · IET Computer Vision · doi:10.1049/iet-cvi.2019.0171 · Cited by 33
“The rotation-invariant layer and a Fisher discriminative layer were added to the existing object detection CNN to solve the issues of rotation variations, within-class diversity, and between-class similarity in [31, 32].”
MethodPutro, M. D., Kurnianggoro, L., & Jo, K.-H. (2020) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2020.3022501 · Cited by 28
“Even objects with varying positions can be resolved by considering rotation invariance and discriminating criteria without changing the existing network architectures [5].”
BackgroundDensely Semantic Enhancement for Domain Adaptive Region-Free Detectors
Zhang, B., Chen, T., Wang, B., Wu, X., Zhang, L., & Fan, J. (2021) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2021.3069034 · Cited by 31
“Besides, explicitly embedding the rotation invariant regularization and the Fisher discrimination criterion into the objective functions of detection framework is proposed to learn more powerful representations [23].”
MethodZero-Shot Aerial Object Detection with Visual Description Regularization
Zang, Z., Lin, C., Tang, C., Wang, T., & Lv, J. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i7.28518 · Cited by 39
He, Y., Chen, W., Wang, S., Liu, T., & Wang, M. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3459589 · Cited by 37
Small Manhole Cover Detection in Remote Sensing Imagery with Deep Convolutional Neural Networks
Liu, W., Cheng, D., Yin, P., Yang, M., Li, E., Xie, M., & Zhang, L. (2019) · ISPRS International Journal of Geo-Information · doi:10.3390/ijgi8010049 · Cited by 27
Capsule Networks for Object Detection in UAV Imagery
Mekhalfi, M. L., Bejiga, M. B., Soresina, D., Melgani, F., & Demir, B. (2019) · Remote Sensing · doi:10.3390/rs11141694 · Cited by 26
Cao, Y., Wang, Y., Peng, J., Zhang, L., Xu, L., Yan, K., & Li, L. (2020) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.2991545 · Cited by 27
“In [20], rotation-invariant regularization and Fisher discrimination regularization were imposed on the CNN features.”
BackgroundMethodYin, H., Wei, Y., Liu, H., Liu, S., Liu, C., & Gao, Y. (2020) · Complexity · doi:10.1155/2020/6843869 · Cited by 26
Enhancing Convolutional Neural Network using Hu’s Moments
Aburass, S., Huneiti, A., & Al-Zoubi, M. B. (2020) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2020.0111216 · Cited by 30
“Although CNN is translation and slight rotation equivariance, it is not translation, scaling or rotation invariant [5],[7],[24],[26],[32],[31].”
BackgroundMargin-Based Adversarial Joint Alignment Domain Adaptation
Zuo, Y., Yao, H., Zhuang, L., & Xu, C. (2021) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2021.3081729 · Cited by 27
“RIFD-CNN [42] adds a rotation-invariant and Fisher discrimination regularizer to achieve rotation-invariance, small within-class scatter and large between-class separation for object detection.”
MethodDermoscopic image retrieval based on rotation-invariance deep hashing
Zhang, Y., Xie, F., Song, X., Zheng, Y., Liu, J., & Wang, J. (2021) · Medical Image Analysis · doi:10.1016/j.media.2021.102301 · Cited by 27
Data augmentation with Mobius transformations
Zhou, S., Zhang, J., Jiang, H., Lundh, T., & Ng, A. Y. (2020) · Machine Learning Science and Technology · doi:10.1088/2632-2153/abd615 · Cited by 24
“Finally, there is adjacent work that follows the aim of data augmentation methods to learn invariant representations, and directly learns network architectures that are invariant to select transformations, such as rotation invariance [22, 23].”
Machine learning paradigms in high-resolution remote sensing image interpretation
Zhou, P., Gong, C., Yao, X., & Junwei, H. (2021) · National Remote Sensing Bulletin · doi:10.11834/jrs.20210164 · Cited by 33
Self-citationArbitrary-angle bounding box based location for object detection in remote sensing image
Sun, F., Li, H., Liu, Z., Li, X., & Wu, Z. (2021) · European Journal of Remote Sensing · doi:10.1080/22797254.2021.1880975 · Cited by 24
Li, H., Hu, C., Zhong, X., Zeng, C., & Shen, H. (2022) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2022.3224555 · Cited by 24
“Such methods include the region based convolutional neural network (R-CNN) [20], fast R-CNN [29], faster R-CNN [30], the feature pyramid network (FPN) [31], and mask R-CNN [32].”
BackgroundInfluentialHan, M., Su, Z., & Na, X. (2023) · Stochastic Environmental Research and Risk Assessment · doi:10.1007/s00477-023-02405-4 · Cited by 24
“Typical applications include applications in image recognition and processing (Cheng et al. 2019), natural language processing (Otter et al. 2021), and processing of high-dimensional chaotic systems that exist in nature (Li et al. 2012).”
BackgroundVehicle-Type Recognition Method for Images Based on Improved Faster R-CNN Model
Bai, T., Luo, J., Zhou, S., Lü, Y., & Wang, Y. (2024) · Sensors · doi:10.3390/s24082650 · Cited by 25
RingMoGPT: A Unified Remote Sensing Foundation Model for Vision, Language, and Grounded Tasks
Wang, P., Hu, H., Tong, B., Zhang, Z., Yao, F., Feng, Y., Zhu, Z., Chang, H., Diao, W., Ye, Q., et al. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3510833 · Cited by 62
A more compact object detector head network with feature enhancement and relational reasoning
Zhang, W., Fu, C., Chang, X., Zhao, T., Li, X., & Sham, C. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.05.034 · Cited by 24
Learning an Invariant and Equivariant Network for Weakly Supervised Object Detection
Feng, X., Yao, X., Shen, H., Cheng, G., Xiao, B., & Han, J. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3275142 · Cited by 25
Self-citationUnsupervised Out-of-Distribution Object Detection via PCA-Driven Dynamic Prototype Enhancement
Wu, A., Deng, C. D., & Liu, W. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3378464 · Cited by 27
“O BJECT detection [1], [2], [3], [4], [5], [6] based on deep neural networks has achieved remarkable success in known categories that are used for training.”
BackgroundRotating-YOLO: A novel YOLO model for remote sensing rotating object detection
Liu, Z., Chen, Y.-K., & Gao, Y. (2024) · Image and Vision Computing · doi:10.1016/j.imavis.2024.105397 · Cited by 25
Simultaneous Subspace Clustering and Cluster Number Estimating Based on Triplet Relationship
Liang, J., Yang, J., Cheng, M.-M., Rosin, P. L., & Wang, L. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2903294 · Cited by 23
Zhang, Z., Jiang, R., Mei, S., Zhang, S., & Zhang, Y. (2019) · IEEE Access · doi:10.1109/access.2019.2960931 · Cited by 23
Guo, W., Li, W., Li, Z., Gong, W., Cui, J., & Wang, X. (2020) · Remote Sensing · doi:10.3390/rs12223750 · Cited by 23
Garajeh, M. K., Weng, Q., Haghi, V. H., Li, Z., Garajeh, A. K., & Salmani, B. (2022) · Canadian Journal of Remote Sensing · doi:10.1080/07038992.2022.2072277 · Cited by 25
“The Adam function has a better performance in practical applications and can correct existing problems in other optimization functions such as large fluctuations in loss/cost function generated by parameter updates with high variance and learning rate vanishing (Gong et al. 2019).”
BackgroundMo, N., Li, Y., Zhu, R., & Xie, H. T. (2019) · Remote Sensing · doi:10.3390/rs11030272 · Cited by 27
Lindeberg, T. (2019) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-019-00915-x · Cited by 28
VSAI: A Multi-View Dataset for Vehicle Detection in Complex Scenarios Using Aerial Images
Wang, J., Teng, X., Li, Z., Yu, Q., Bian, Y., & Wei, J. (2022) · Drones · doi:10.3390/drones6070161 · Cited by 32
Classification of Transformed and Geometrically Distorted Images using Convolutional Neural Network
Aburass, S., Huneiti, A., & Al-Zoubi, M. B. (2022) · Journal of Computer Science · doi:10.3844/jcssp.2022.757.769 · Cited by 21
“758 equivariance, it is not a translation, scaling or rotation invariant (Cohen et al., 2017; Hinton et al., 2012; Worrall et al., 2017; Cheng et al., 2018; Aburass et al., 2020).”
BackgroundMethodContext Aware Edge-Enhanced GAN for Remote Sensing Image Super-Resolution
Ren, Z., He, L., & Lu, J. (2023) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2023.3333271 · Cited by 31
“With the development of remote sensing image application technology, remote sensing images are extensively used in hyperspectral application [1], [2], [3], [4], [5], [6], [7], [8], object detection [9], [10], change detection [11], [12], [13], and other fields.”
MethodQuasi Fourier-Mellin Transform for Affine Invariant Features
Yang, J., Lu, Z., Tang, Y. Y., Yuan, Z., & Chen, Y. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.2967578 · Cited by 21
“Furthermore, it has been shown (see [25], [26], etc.”
CUA Loss: Class Uncertainty-Aware Gradient Modulation for Robust Object Detection
Kim, J. U., Kim, S. T., Lee, H. J., Lee, S., & Ro, Y. M. (2020) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2020.3042219 · Cited by 24
Zeng, X., Wei, S., Wei, J., Zhou, Z., Shi, J., Zhang, X., & Fan, F. (2021) · Remote Sensing · doi:10.3390/rs13142788 · Cited by 21
“[2] proposed rotation-invariant and fisher discriminative CNN (RIFD-CNN) to impose CNN features with fisher discriminative regularizer and rotation-invariant regularizer.”
MethodR²IPoints: Pursuing Rotation-Insensitive Point Representation for Aerial Object Detection
Yao, X., Shen, H., Feng, X., Cheng, G., & Han, J. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3173373 · Cited by 22
“For instance, Cheng et al. [29] introduced a rotation-invariant regularizer and a Fisher discrimination regularizer on CNN features to deal with the challenges of object rotation, within-class diversity, and between-class similarity in aerial images.”
BackgroundSelf-citationMultigrained Angle Representation for Remote-Sensing Object Detection
Wang, H., Huang, Z., Chen, Z., Song, Y., & Li, W. (2022) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2022.3212592 · Cited by 29
“Therefore, to balance accuracy, speed, and parameter stability, the selection range of Cθ is restricted to [3, 4, 5] in the proposed MGAR.”
BackgroundMethodInfluentialOn Single-Model Transferable Targeted Attacks: A Closer Look at Decision-Level Optimization
Sun, X., Cheng, G., Li, H., Pei, L., & Han, J.-W. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3276331 · Cited by 22
“G IVEN the revealed security vulnerability (dubbed adversarial vulnerability) [1], [2] of Deep Neural Networks (DNNs), the wide applications of DNNs in image understanding [3], [4], [5], [6], [7], [8], [9], [10] suffer from a potential safety concern.”
BackgroundSelf-citationDeep built-structure counting in satellite imagery using attention based re-weighting
Shakeel, A., Sultani, W., & Ali, M. (2019) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2019.03.014 · Cited by 21
Metadata
- Authors
- Cheng, G., Han, J., Zhou, P., & Xu, D.
- Affiliations
- Northwestern Polytechnical University (CN)
- The University of Sydney (AU)
- Venue
- IEEE Transactions on Image Processing
- Published
- 2019-01
- Volume
- 28
- Issue
- 1
- Pages
- 265-278
- ISSN
- 1057-7149 (ISSN-L)
- 1941-0042
- OpenAlex
- W2888527098
- Cited by
- 408
- References
- 77
- Citations by year
- 2026: 12 · 2025: 19 · 2024: 26 · 2023: 46 · 2022: 64 · 2021: 77 · 2020: 77 · 2019: 86
- FWCI
- 30.15
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Image Processing (publishedVersion)
- PubMed (publishedVersion)
- Funders
- National Natural Science Foundation of China (61772425, 61473231, 61790552)
- Natural Science Basic Research Plan in Shaanxi Province of China (2017JM6044, 2018KJXX-029)
- National Aerospace Science Foundation of China (2017ZC53032)
- Fundamental Research Funds for the Central Universities (3102018zy023)
- Australian Research Council (FT180100116)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- object detection, rotation invariance, within-class scatter, Faster R-CNN, R-FCN, aerial images
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.