Learning Rotation-Invariant Convolutional Neural Networks for Object Detection in VHR Optical Remote Sensing Images

By Gong Cheng, Peicheng Zhou, Junwei Han. 2016

1,611 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Remote Sensing Image Scene Classification: Benchmark and State of the Art (2017, background): “models have shown impressive feature representation capability for a wide range of applications including remote sensing image scene classification [13, 17, 45, 50, 82, 143-158].”
Why do deep convolutional networks generalize so poorly to small image transformations? (2018, methodology): “Alternatively, one could use specially designed features such as SIFT in which invariance is hard coded [7, 29, 30, 31] or neural network architectures that explicitly enforce invariance [36, 19, 5, 6, 11, 12, 47, 46, 8].”
Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community (2017, methodology): “Cheng et al.1 discuss traditional hand-crafted features such as the histogram of ordered gradients (HOG), which is a feature of the scale-invariant feature transform (SIFT), color histograms, local binary patterns (LBP), etc.”
Building Instance Classification Using Street View Images (2018, background): “For e ﬀ ectively dealing with the problem of object rotation variations, a rotation-invariant CNN model is proposed in [47].”
Self-Attention for Raw Optical Satellite Time Series Classification (2019, methodology): “Rotation-invariant feature extraction [51] for scene classification [52] and metric learning [53] has been similarly tested for remote sensing image classification.”
Deep Learning for Geophysics: Current and Future Trends (2021, background): “…rotation variations, a rotation-invariant CNN for object detection in very high-resolution optical remote sensing images was proposed, where a rotation-invariant layer was introduced by enforcing the training samples before and after rotation to share the same features (Cheng et al., 2016).”
Learning a Rotation Invariant Detector with Rotatable Bounding Box (2017, methodology): “A CNN based pio-neering study upon vehicle detection in satellite images is presented in[2], followed by many approaches that focus on different types of geospatial object detection problems [14, 20, 21, 5, 26, 17].”
Learning RoI Transformer for Oriented Object Detection in Aerial Images (2019, methodology): “Many of recent progress on object detection in aerial images have beneﬁted a lot from the R-CNN frameworks [9, 8, 32, 2, 29, 38, 6, 12, 16].”
Object Detection in 20 Years: A Survey (2019, methodology): “After the great success of RCNN in 2014, deep CNN has been soon applied to remote sensing target detection [275, 276, 400, 401].”
A Spatial-Temporal Attention-Based Method and a New Dataset for Remote Sensing Image Change Detection (2020, methodology): “In light of the good performance in computer vision, deep-CNN-based methods have been widely applied in remote sensing tasks [25], such as land-use classification [39], semantic segmentation [40,41], image super-resolution [42], object detection [43,44] and change detection [5].”
SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated Objects (2018, background): “The public benchmark NWPU VHR-10 [4] contains 10-class geospatial object for detection.”
Algorithms for semantic segmentation of multispectral remote sensing imagery using deep learning (2017, background): “Deep learning approaches to classify and analyze RGB remote sensing imagery have advanced considerably thanks to deep learning, including the use of DCNNs pre-trained on ImageNet and unsupervised feature extraction [17, 18, 19].”
Deep learning in remote sensing: a review (2017, background): “[76] proposed an effective approach to learning a rotation-invariant CNN (RICNN) to improve invariance to object rotation.”
Transformers in Remote Sensing: A Survey (2022, methodology): “GANsformer [40] Detection DIOR [83], NWPU VHR-10 [85] mAP Introduces an efficient transformer, with reduced parameters, as a branch network to capture global features along with a generative model to expand the input image ahead of backbone.”
Nested Network With Two-Stream Pyramid for Salient Object Detection in Optical Remote Sensing Images (2019, methodology): “databases, such as NWPU VHR-10 data set [38], AID data set [39], LEVIR data set [40], and NWPU-RESISC45 data set [41].”
Land cover mapping at very high resolution with rotation equivariant CNNs: towards small yet accurate models (2018, background): “We call this property equivariance, and it is a property that recently attracted a lot of interest in image analysis (Lei et al., 2012; Cheng et al., 2016).”