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 201–300 of 408
Object Detection Based on Efficient Multiscale Auto-Inference in Remote Sensing Images
Zhang, S., Mu, X., Kou, G., & Zhao, J. (2020) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2020.3004061 · Cited by 20
An Improved FBPN-Based Detection Network for Vehicles in Aerial Images
Wang, B., & Gu, Y. (2020) · Sensors · doi:10.3390/s20174709 · Cited by 20
“Many papers have revealed that the netw rk depth is important for performance improvement, and some nontrivial visual detection tasks have also greatly benefited from very deep models [23,24].”
BackgroundMethodOLCN: An Optimized Low Coupling Network for Small Objects Detection
Yuan, Y., & Zhang, Y. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3122190 · Cited by 23
CODrone: A Comprehensive Oriented Object Detection Benchmark for UAV
Ye, K., Tang, H., Liu, B., Dai, P., & Cao, L. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3677376 · Cited by 27
“Meanwhile, NWPU VHR-10 [25] contains 800 very-high-resolution images covering ten geospatial object categories.”
Wang, H., Liao, Y., Li, Y., Fang, Y., Ni, S., Luo, Y., & Jiang, B. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3221101 · Cited by 19
“To manage diverse object orientations, Fu et al. [30] expanded training samples from multiple angles through data augmentation, and Cheng et al. [31] introduced rotation-insensitive submodules.”
BackgroundLearning Remote Sensing Object Detection With Single Point Supervision
He, S., Zou, H., Wang, Y., Li, B., Cao, X., & Jing, N. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3343806 · Cited by 28
Cyber-physical Support of Daily Activities: A Robot/Smart Home Partnership
Pereyda, C., Raghunath, N., Minor, B., Wilson, G., Schmitter‐Edgecombe, M., & Cook, D. J. (2019) · ACM Transactions on Cyber-Physical Systems · doi:10.1145/3365225 · Cited by 22
“We employ a convolutional neural network (CNN) that has demonstrated success for general object detection [60].”
MethodScale-Invariant Scale-Channel Networks: Deep Networks That Generalise to Previously Unseen Scales
Jansson, Y., & Lindeberg, T. (2022) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-022-01082-2 · Cited by 32
Eagle-Eye-Inspired Attention for Object Detection in Remote Sensing
Liu, K., Huang, J., & Li, X. (2022) · Remote Sensing · doi:10.3390/rs14071743 · Cited by 19
“Some domain algorithms of remote sensing are RICAOD [77], RIFD-CNN [31], RICNN-finetuning [30], HRCNN-regression [72], and FRCNN TC [73].”
BackgroundMethodNIRNet: Noise Incentive Robust Network in Remote Sensing Object Detection Under Cloud Corruption
Zhang, P., Cheng, G., Lang, C., Xie, X., & Han, J. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3581342 · Cited by 32
“Plentiful studies contribute to feasible solutions for recognizing objects with arbitrary orientations in remote sensing object detection [46], [47], [48].”
Self-citationASPCNet: Deep adaptive spatial pattern capsule network for hyperspectral image classification
Wang, J., Tan, X., Lai, J., & Li, J. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.02.058 · Cited by 18
Sampling Equivariant Self-Attention Networks for Object Detection in Aerial Images
Yang, G.-Y., Li, X.-L., Xiao, Z.-K., Mu, T.-J., Martin, R. R., & Hu, S.-M. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3327586 · Cited by 19
“Xie et al. [48] present an effective two-stage oriented object detection framework, Oriented R-CNN, with an oriented region proposal network and an oriented R-CNN head.”
BackgroundMultiscale Feature Adaptive Fusion for Object Detection in Optical Remote Sensing Images
Lv, H., Qian, W., Chen, T., Yang, H., & Zhou, X. (2022) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2022.3178787 · Cited by 19
Knowledge Amalgamation for Object Detection With Transformers
Zhang, H., Mao, F., Xue, M., Fang, G., Feng, Z., Song, J., & Song, M. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3263105 · Cited by 23
“[50], and oriented detectors [51], [52], [53], [54].”
Scale Optimization Using Evolutionary Reinforcement Learning for Object Detection on Drone Imagery
Zhang, J., Yang, X., He, W., Ren, J., Zhang, Q., Zhao, Y., Bai, R., He, X., & Liu, J. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i1.27795 · Cited by 29
A Survey on Deep Domain Adaptation and Tiny Object Detection Challenges, Techniques and DatasetsPreprint
Muzammul, M., & Li, X. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.07927 · Cited by 27
“In (Cheng et al., 2018), the Rotation-Invariant and Fisher Discriminative CNN (RIFD) was proposed to include a rotation-invariant regularizer and a fisher discrimination regularizer on multi-scale features from CNN[556].”
MethodProbabilistic Selective Encryption of Convolutional Neural Networks for Hierarchical Services
Tian, J., Zhou, J., & Duan, J. (2021) · 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/cvpr46437.2021.00224 · Cited by 25
“Convolutional Neural Network (CNN) has been used to achieve unparalleled results across various tasks such as object detection [5, 10, 20], super-resolution reconstruction [9, 13, 15], and image inpainting [36, 37, 38].”
MethodRotation-Invariant Deep Embedding for Remote Sensing Images
Kang, J. S., Fernández-Beltrán, R., Wang, Z., Sun, X., Ni, J., & Plaza, A. (2021) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2021.3088398 · Cited by 18
“[44] proposed an effective method to train rotation-invariant and Fisher discriminative CNN models for object detection purposes.”
MethodTowards Rotation Invariance in Object Detection
Kalra, A., Stoppi, G., Brown, B., Agarwal, R., & Kadambi, A. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00351 · Cited by 17
“Several methods like [12, 8, 25], aim to achieve rotation invariance when predicting rotated bounding boxes.”
MethodChen, Y., Liu, Q., Wang, T., Wang, B., & Meng, X. (2021) · Remote Sensing · doi:10.3390/rs13214386 · Cited by 18
“In the recent research work of remote sensing object detection [33,34], rotation invariance can significantly improve the detection effect for this problem.”
DGRNet: A Dual-Level Graph Relation Network for Video Object Detection
Qi, Q., Hou, T., Lu, Y., Yan, Y., & Wang, H. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3285136 · Cited by 19
“Recently, some variants [40], [41], [42] of R-CNN have been proposed to further improve performance.”
BackgroundJiang, R., Mei, S., Ma, M., & Zhang, S. (2020) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2020.3021283 · Cited by 17
“2) Comparison With State-of-the-Arts: Our proposed NSS is compared with four baselines, including Ti-Pooling [19], RICNN [1], [18], iCNN [25], Double-Net [38], and P4CNN [40], and the comparison results are shown in Table VI”
BackgroundMethodInfluentialMulti-category classification with label noise by robust binary loss
Liu, D., Zhao, J., Wu, J., Yang, G., & Lv, F. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.01.031 · Cited by 19
Choi, E., Chae, S., & Kim, J. (2019) · Sensors · doi:10.3390/s19194218 · Cited by 17
“In this event, we can consider rotation-invariant methods [22,23].”
MethodRepresentation and Classification of Auroral Images Based on Convolutional Neural Networks
Yang, Q., & Zhou, P. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.2969245 · Cited by 16
Self-citationA Study on Classification and Detection of Small Moths Using CNN Model
Hilal, A. M., Alsolai, H., Al‐Wesabi, F. N., Al-Hagery, M. A., Hamza, M. A., & Duhayyim, M. A. (2021) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2022.022554 · Cited by 18
In Your Face: Person Identification Through Ratios and Distances Between Facial Features
Alsawwaf, M., Chaczko, Z., Kulbacki, M., & Sarathy, N. (2021) · Vietnam Journal of Computer Science · doi:10.1142/s2196888822500105 · Cited by 17
Wang, Y., Luo, Z., & Luo, J. (2023) · The Science of The Total Environment · doi:10.1016/j.scitotenv.2023.166506 · Cited by 16
A review on deep learning-based object tracking methods
Uke, N., Futane, P. R., Deshpande, N., & Uke, S. N. (2024) · Multiagent and Grid Systems · doi:10.3233/mgs-230126 · Cited by 15
Adaptive multiscale feature for object detection
Yu, X.-Y., Wu, S., Lu, X., & Gao, G. (2021) · Neurocomputing · doi:10.1016/j.neucom.2021.04.002 · Cited by 16
Multiscale Conditional Relationship Graph Network for Referring Relationships in Images
Zhu, J., & Wang, H. (2021) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2021.3079278 · Cited by 17
“Unfortunately, it is challenging for object detectors [7], [8] to directly detect entities in visual relationship benchmark data sets.”
BackgroundLearning Rotated Inscribed Ellipse for Oriented Object Detection in Remote Sensing Images
He, X., Ma, S., He, L., Ru, L., & Wang, C. (2021) · Remote Sensing · doi:10.3390/rs13183622 · Cited by 16
“The method in [24,25] used a rotation-invariant convolutional neural network to address the problem of inter-class similarity and intra-class diversity in multi-class RSI object detection.”
MethodFast template matching and object detection techniques using φ-correlation and binary circuits
Annaby, M. H., & Fouda, Y. M. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15564-x · Cited by 16
“Template recognition, matching, and tracking as well as object detection are major signal and image processing techniques that have many applications in different disciplines, see e.g. [6, 17, 27, 39].”
Scale-Recursive Network with point supervision for crowd scene analysis
Dong, Z., Zhang, R., Shao, X., & Li, Y. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.12.070 · Cited by 16
“Furthermore, recent studies [17,18] try to learn more useful CNN features that are rotation invariant, which is an important component in object detection methods.”
BackgroundImproving Multi-Person Pose Tracking With a Confidence Network
Fu, Z., Zuo, W., Hu, Z., Liu, Q., & Wang, Y. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3330532 · Cited by 17
Boosting Semi-Supervised Object Detection in Remote Sensing Images With Active Teaching
Zhang, B., Wang, Z., & Du, B. (2024) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2024.3357098 · Cited by 20
“Additionally, remote sensing images often contain objects with various orientations and scales [2], [3], [4], [5], [6], which further complicates the labeling process.”
BackgroundRotation and Translation Invariant Palmprint Recognition With Biologically Inspired Transform
Zhou, X., Zhou, K., & Shen, L. (2020) · IEEE Access · doi:10.1109/access.2020.2990736 · Cited by 14
“The phase congruency function in terms of the Fourier series expansion of a signal at some location x is defined as the following equation [60], where A n is the amplitude of the nth Fourier component, and ϕ n ( x ) denotes the local phase of the Fourier component at The maximum phase congruency can…”
BackgroundMethodHuang, W., Li, G., Jin, B., Chen, Q., Yin, J., & Huang, L. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3135935 · Cited by 15
Object detection for panoramic images based on MS‐RPN structure in traffic road scenes
Tong, G., Chen, H., Li, Y., Du, X., & Zhang, Q. (2019) · IET Computer Vision · doi:10.1049/iet-cvi.2018.5304 · Cited by 14
“In addition, rotation-invariant and Fisher discriminative CNN (RIFD-CNN) [23] can learn much richer CNN high-level representation features for object detection by explicitly imposing a rotation-invariant regulariser and a Fisher discrimination regulariser on the CNN features.”
BackgroundYang, F., Wang, Z., & Heinze‐Deml, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.11235 · Cited by 31
“These approaches are based on augmenting the training data either statically [17, 30, 7, 45], ie.”
BackgroundMethodAdaptive Discrepancy Masked Distillation for remote sensing object detection
Li, C., Cheng, G., & Han, J. (2025) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2025.02.006 · Cited by 13
Self-citationObject Detection in Convolution Neural Networks Using Iterative Refinements
Aroulanandam, V. V., Latchoumi, T. P., Bhavya, B., & Sultana, S. F. (2019) · Revue d intelligence artificielle · doi:10.18280/ria.330506 · Cited by 15
“These approaches require re-training and offline learning for the training when applied to the separate datasets [11-13].”
MethodFisher Regularized ε-Dragging for Image Classification
Chen, Z., Wu, X., & Kittler, J. (2022) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2022.3175008 · Cited by 13
“Besides, recently some deep Fisher-like discriminative classification networks have been proposed to solve the problem of object detection [5] and remote sensing image”
BackgroundCOVID-19 Infection Segmentation and Severity Assessment Using a Self-Supervised Learning Approach
Song, Y., Liu, J., Liu, X., & Tang, J. (2022) · Diagnostics · doi:10.3390/diagnostics12081805 · Cited by 15
Sirirattanapol, C., Nagai, M., Witayangkurn, A., Pravinvongvuth, S., & Ekpanyapong, M. (2019) · ISPRS International Journal of Geo-Information · doi:10.3390/ijgi8030128 · Cited by 13
Liu, Y., Gong, X., & Yang, Y. (2020) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2020.2992816 · Cited by 13
“In order to prevent the network from being interfered by various rotating low-altitude remote sensing images, we are inspired by Cheng et al. [11] and Gong et al. [12], [13] to make the trained network have the characteristic of rotation-invariant.”
MethodYan, H. (2020) · Remote Sensing Letters · doi:10.1080/2150704x.2020.1770364 · Cited by 14
“Because of CNN -based methods ’ powerful capability on features representation, some CNN -based approaches have been proposed for aircraft detection (Zhang et al. 2018; Yang et al. 2018; Cai et al. 2018; Zhang et al. 2019; Cheng et al. 2018).”
MethodU-Net Based Road Area Guidance for Crosswalks Detection from Remote Sensing Images
Chen, Z., Luo, R., Li, J., Du, J., & Wang, C. (2021) · Canadian Journal of Remote Sensing · doi:10.1080/07038992.2021.1894915 · Cited by 14
“…performance, researchers have proposed many methods to deal with those problems in object detection from remote sensing images (Zhang et al. 2018; Cheng et al. 2019; Wang et al. 2019; Li et al. 2018; Zhou et al. 2021; Liu et al. 2017; Li et al. 2019; Chen et al. 2020; Zhang et al. 2019; Yao et…”
BackgroundMethodRMDC: Rotation-mask deformable convolution for object detection in top-view fisheye cameras
Wei, X., Wei, Y., & Lu, X. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.06.116 · Cited by 12
A robust training method for object detectors in remote sensing image
Lin, J., Zhao, Y. F., Wang, S., & Tang, Y. (2023) · Displays · doi:10.1016/j.displa.2023.102618 · Cited by 12
Diffusion-based remote sensing image fusion for classification
Jiang, Y., Liu, S., & Wang, H. (2024) · Applied Intelligence · doi:10.1007/s10489-024-06217-z · Cited by 17
Optimized Projection and Fisher Discriminative Dictionary Learning for EEG Emotion Recognition
Gu, X., Fan, Y., Zhou, J., & Zhu, J. (2021) · Frontiers in Psychology · doi:10.3389/fpsyg.2021.705528 · Cited by 14
“In particular, to promote the discrimination ability of the model, we utilize the model according to the Fisher discrimination criterion (Gong et al., 2019) under the structure of dictionary learning.”
MethodSunwoo, H., Choi, W., Na, S., Kim, C., & Heo, S. (2022) · Applied Sciences · doi:10.3390/app12063136 · Cited by 13
A literature review on remote sensing scene categorization based on convolutional neural networks
Kaul, A., & Kumari, M. (2023) · International Journal of Remote Sensing · doi:10.1080/01431161.2023.2204200 · Cited by 12
“Furthermore, in addition to reducing the cross-entropy loss, Cheng et al. (2018) imposes another type of regularization technique, namely metric learning regularization, which results in the extraction of more discriminative features.”
BackgroundMethodNi, T., Ni, Y., Xue, J., & Wang, S. (2021) · Frontiers in Psychology · doi:10.3389/fpsyg.2021.721266 · Cited by 13
“Following Gong et al. (2019) and Ma et al. (2011), the PCA criterion is used to preserve discriminative knowledge in the subspace.”
MethodA feature fusion deep-projection convolution neural network for vehicle detection in aerial images
Wang, B., & Xu, B. (2021) · PLoS ONE · doi:10.1371/journal.pone.0250782 · Cited by 11
“Cheng et al tried to solve the detecting problem related to object rotation, within-class diversity and inter-class similarity by adding a rotation-invariant regularizer and a fisher discrimination regularizer to the existing neural networks in 2018 [25].”
BackgroundFuertes, D., del‐Blanco, C. R., Carballeira, P., Jaureguizar, F., & Garcia, N. N. (2022) · Digital Signal Processing · doi:10.1016/j.dsp.2022.103473 · Cited by 12
Jibao, L., Kang, X., Xukun, L. U., & Li, S. (2022) · National Remote Sensing Bulletin · doi:10.11834/jrs.20221555 · Cited by 16
Scene Classification of High Resolution Remote Sensing Images Via Self-Paced Deep Learning
Yao, X., Yang, L., Cheng, G., Han, J., & Guo, L. (2019) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss.2019.8898387 · Cited by 11
“Recently, deep learning have demonstrated the superiority of learning powerful features [1, 13-15] and have achieved very promising results for HRRS scene classification [9, 12, 15-18].”
BackgroundSelf-citationJoint Learning of NNeXtVLAD, CNN and Context Gating for Micro-Video Venue Classification
Liu, W., Huang, X., Cao, G., Zhang, J., Song, G., & Yang, L. (2019) · IEEE Access · doi:10.1109/access.2019.2922430 · Cited by 11
“Recently, metric learning [31] and fisher discriminative CNN [32], [33] have been proposed to enforce CNN features to be more discriminative.”
BackgroundDetection of power line insulators on digital images with the use of laser spots
Tomaszewski, M., Michalski, P., Ruszczak, B., & Zator, S. (2019) · IET Image Processing · doi:10.1049/iet-ipr.2018.6284 · Cited by 11
VLPRSDet: A vision–language pretrained model for remote sensing object detection
Liu, D., Liang, X., Qi, Y., Xi, Y., Jin, J., & Zhang, J. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131712 · Cited by 15
SRF: SpectrumRecombineFormer for Hyperspectral Image Classification
Jing, W., Kang, P., Di, D., Gu, J., Li, L., Emam, M., Mohaisen, L., Yang, X., & Li, C. (2025) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3715698 · Cited by 12
“Without the SRC and CF modules, the Classical Transformer (ViT) achieves the lowest classification accuracy, followed by the Swin-Transformer.”
On-device facial verification using NUF-Net model of deep learning
Termritthikun, C., Jamtsho, Y., & Muneesawang, P. (2019) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2019.07.018 · Cited by 10
Qiu, T., He, H., Liang, X., Chen, F., Chen, Z., & Liu, Y. (2022) · Frontiers in Environmental Science · doi:10.3389/fenvs.2022.981486 · Cited by 10
“Recently, CNNs have been applied to remote sensing image classi fi cation and achieved good results (Scott et al., 2017; Cheng et al., 2018a; Cheng et al., 2018b; Weng et al., 2018; Zhou et al., 2019).”
BackgroundAdaptive dynamic networks for object detection in aerial images
Wu, Z., & Yan, H. (2022) · Pattern Recognition Letters · doi:10.1016/j.patrec.2022.12.022 · Cited by 10
Multitask Learning Mechanism for Remote Sensing Image Motion Deblurring
Fang, J., Cao, X., Wang, D., & Xu, S. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.3047636 · Cited by 11
“Recently, deep neural networks especially CNNs have achieved significant performances in many tasks such as image recognition [42], [43], object detection [44], semantic segmentation [45], etc.”
BackgroundImage quality assessment for intelligent emergency application based on deep neural network
Wei, G., & Sheng, Z. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102581 · Cited by 10
Objects and scenes classification with selective use of central and peripheral image content
Alameer, A., Degenaar, P. A., & Nazarpour, K. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102698 · Cited by 10
“Recent advances in object detection involved introducing a simple method to train rotation-invariant and Fisher discriminative CNN models [16] to boost CNN performance.”
MethodDAE-YOLO: Remote Sensing Small Object Detection Method Integrating YOLO and State Space Models
Li, B., Kang, Y., Ding, Y., Li, S., Zhang, Z., & Ma, D. (2025) · Remote Sensing · doi:10.3390/rs18010109 · Cited by 11
Adaptive image zoom-in with bounding box transformation for UAV object detection
Wang, T., Lin, C.-Y., Tang, C., Zhou, J., Xiong, D., Li, J., Zhao, J., & Lv, J. (2026) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2026.01.036 · Cited by 14
“In addition to improving standard bounding box detection accuracy as in normal object detection setting, some works study how to predict rotated bounding boxes to better fit the aerial objects [53, 54, 55].”
Zhou, Y. Z., Xia, L., Zhao, J., Yao, R., & Liu, B. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15608-2 · Cited by 15
Yao, X., Feng, X., Cheng, G., Han, J., & Guo, L. (2019) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss.2019.8899285 · Cited by 9
“More recently, deep learning algorithms especially CNNs have achieved a great success for various applications in the field of nature images [9-17].”
BackgroundSelf-citationA multi-branch separable convolution neural network for pedestrian attribute recognition
Junejo, I. N., & Ahmed, N. (2020) · Heliyon · doi:10.1016/j.heliyon.2020.e03563 · Cited by 9
“Some recent works focus on lowdimensional feature mappings, or imposing regularization constraint to learn rotation-invariant features as well [20, 21, 22, 23].”
BackgroundNovel Hand Gesture Alert System
Mambou, S., Krejcar, O., Marešová, P., Selamat, A., & Kuča, K. (2019) · Applied Sciences · doi:10.3390/app9163419 · Cited by 9
“We anticipate that in our future work, we will associate the facial expression with the alert decision in order to investigate ResNext [48–51] or a faster impact of the CNN-Region mode, at the same time as the detector and classifier [52,53].”
MethodShape-based filter for micro-aneurysm detection
Zhang, X., Xiao, Z., Zhang, F., Ogunbona, P., Xi, J., & Tong, J. (2020) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2020.106620 · Cited by 9
Liu, W., He, C., & Sun, L. (2021) · Remote Sensing · doi:10.3390/rs13163196 · Cited by 8
Esfahani, M. A., Wang, H., Bashari, B., Wu, K., & Yuan, S. (2021) · Applied Soft Computing · doi:10.1016/j.asoc.2021.107424 · Cited by 11
ULN: An efficient face recognition method for person wearing a mask
Lu, H., & Zhuang, Z. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-13495-7 · Cited by 9
Geng, J., Xu, Z., Zhao, Z., & Jiang, W. R. (2022) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2022.3207382 · Cited by 9
“[4] proposed a rotation-invariant and Fisher 36 discriminative CNN model, which can reduce the influence of 37”
BackgroundTian, S., Cao, L., Kang, L., Xing, X., Tian, J., Du, K., Sun, K., Fan, C., Fu, Y., & Zhang, Y. (2022) · Remote Sensing · doi:10.3390/rs14194951 · Cited by 8
Zhang, S., Fang, G., Song, J., Li, R., Gu, Y., Lin, Z., Zhou, C.-C., Dai, Y., & Kong, X. (2024) · Research in Astronomy and Astrophysics · doi:10.1088/1674-4527/ad6fe6 · Cited by 9
Small object segmentation with fully convolutional network based on overlapping domain decomposition
Park, J., Kwon, D., Choi, B. W., Kim, G. Y., Kim, K., & Kwon, J. (2019) · Machine Vision and Applications · doi:10.1007/s00138-019-01023-x · Cited by 8
YADA: you always dream again for better object detection
Nguyen, K.-D., Nguyen, K., Le, D.-D., Duong, D. A., & Nguyen, T. (2019) · Multimedia Tools and Applications · doi:10.1007/s11042-019-07888-4 · Cited by 8
Zhou, W., Shen, J., Liu, N., Xia, S., & Sun, H. (2022) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2022.3202186 · Cited by 9
“cle (UAV) technology, high-resolution aerial images 27 are becoming more and more easily available [1].”
BackgroundZeng, Z., Liu, H., Chen, F., & Tan, X. (2023) · Remote Sensing · doi:10.3390/rs15092393 · Cited by 8
Pedestrian detection using multi-channel visual feature fusion by learning deep quality model
Yu, P., Zhao, Y. J., Zhang, J., & Xie, X. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102579 · Cited by 8
Diverse receptive field network with context aggregation for fast object detection
Xie, S., Liu, C., Gao, J., Li, X., Luo, J., Fan, B., Chen, J., Pu, H., & Peng, Y. (2020) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2020.102770 · Cited by 9
Li, Z., You, Y., & Liu, F. (2020) · IEEE Access · doi:10.1109/access.2020.3033469 · Cited by 8
“Reference [21] introduced a rotation-invariant regularizer and fisher discrimination regularizer to construct a new objective function to improve the performance of object detection.”
BackgroundWeakly Supervised Part-Based Method for Combined Object Detection in Remote Sensing Imagery
Qian, W., Yan, Z., Zhu, Z., & Yin, W. (2022) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2022.3179026 · Cited by 10
“Furthermore, in their more recent work, Cheng et al. [9] proposed the rotation-invariant and Fisher discriminative CNN (RIFD-CNN) model by imposing a rotation-invariant regularizer and a Fisher discrimination regularizer on the CNN features.”
MethodAn Extensive Review of Deep Learning Driven Remote Sensing Image Classification Models
Rathod, V. V., Rana, D. P., & Mehta, R. G. (2022) · International Conference on Intelligent Computing, Instrumentation and Control Technologies (ICICICT) · doi:10.1109/icicict54557.2022.9917583 · Cited by 8
Heterogeneous domain adaptation with label and structural consistency
Huang, J., Zhou, Z., Shang, J., & Niu, C. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-08731-x · Cited by 9
Wang, Y., Luo, Z., Kong, Y., & Luo, J. (2024) · Sustainability · doi:10.3390/su16114531 · Cited by 7
“The common monitoring parameters”
Zheng, Z., & Cao, J. (2019) · IEEE Access · doi:10.1109/access.2019.2936295 · Cited by 7
“More recently, deep neural networks have drawn lots of attention, which aim at capture higher-level features of the image adaptively [23], [24].”
BackgroundTask-Specific Heterogeneous Network for Object Detection in Aerial Images
Yu, Y., Yang, X., Li, J., & Gao, X. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3311966 · Cited by 13
“O BJECT detection in aerial images is a fundamental and challenging task, involving predicting accurate bounding boxes and corresponding categories of objects within an image [1], [2], [3], [4], [5].”
BackgroundSorting Convolution Operation for Achieving Rotational Invariance
Mo, H., & Zhao, G. (2024) · IEEE Signal Processing Letters · doi:10.1109/lsp.2024.3381909 · Cited by 7
“We do not select STN [14], TI-Pooling [19], and several methods utilizing rotation-invariant loss functions [22], [38] for comparison, because their invariance relies on data augmentation.”
MethodOrtDet: An Orientation Robust Detector via Transformer for Object Detection in Aerial Images
Zhao, L. D., Liu, T., Xie, S., Huang, H., & Qi, J. (2022) · Remote Sensing · doi:10.3390/rs14246329 · Cited by 7
Multiple object tracking with appearance feature prediction and similarity fusion
Li, Z., Chen, J., & Bi, J. (2023) · IEEE Access · doi:10.1109/access.2023.3279868 · Cited by 12
“In 2018, Cheng et al. introduced a model based on a convolutional neural network (CNN) that incorporates a rotation-invariant regularizer and a Fisher discriminant regularizer to improve object detection performance [3].”
MethodCODet: Component Object Detector Extracting Structural Features Based on Target Characteristics
Zhu, Z., Sun, X., Diao, W., Chen, K., He, Q., Xu, G., & Fu, K. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3281331 · Cited by 9
“The feature hierarchy misalignment problem exists in the detection task, especially for component objects (COs) in RS scenes [20], [21], [22], [23], [24].”
BackgroundMobile Communication Base Station Antenna Measurement Using Unmanned Aerial Vehicle
Zhai, Y., Ke, Q., Xu, Y. W., Deng, W., Gan, J., Zeng, J., Zhou, W., Scotti, F., Labati, R. D., & Piuri, V. (2019) · IEEE Access · doi:10.1109/access.2019.2935613 · Cited by 8
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.