Deformable Faster R-CNN with Aggregating Multi-Layer Features for Partially Occluded Object Detection in Optical Remote Sensing Images

Abstract

The region-based convolutional networks have shown their remarkable ability for object detection in optical remote sensing images. However, the standard CNNs are inherently limited to model geometric transformations due to the fixed geometric structures in its building modules. To address this, we introduce a new module named deformable convolution that is integrated into the prevailing Faster R-CNN. By adding 2D offsets to the regular sampling grid in the standard convolution, it learns the augmenting spatial sampling locations in the modules from target tasks without additional supervision. In our work, a deformable Faster R-CNN is constructed by substituting the standard convolution layer with a deformable convolution layer in the last network stage. Besides, top-down and skip connections are adopted to produce a single high-level feature map of a fine resolution, on which the predictions are to be made. To make the model robust to occlusion, a simple yet effective data augmentation technique is proposed for training the convolutional neural network. Experimental results show that our deformable Faster R-CNN improves the mean average precision by a large margin on the SORSI and HRRS dataset.

Arbitrary-Oriented Inshore Ship Detection based on Multi-Scale Feature Fusion and Contextual Pooling on Rotation Region Proposals

“…Object detection frameworks using deep neural networks have been applied to remote sensing applications [33,34] since they were proved to outperform traditional methods in natural images. In the field of inshore ship detection, many researchers have delved into the development of target detection with deep learning methods.…”

Multi-Scale Geospatial Object Detection Based on Shallow-Deep Feature Extraction

“…As a result, the proposed SD-MS framework outperforms all comparison approaches for all ten classes of the NWPU VHR-10 dataset, which demonstrates the superiority of the proposed method compared with the eight other methods. In addition, to quantitatively evaluate the proposed SD-MS model, this study compared it with eight existing methods: rotation-invariant CNN (RICNN) [15], region proposal networks with faster R-CNN (R-P-faster R-CNN) (R-P-F-R-CNN ) [50], deformable R-FCN (D-R-FCN) [51], collection of part detectors (COPD) [11], position-sensitive balancing (PSB) [20], deformable faster R-CNN (D-F-R-CNN) [52], recurrent detection with activated semantics (RDAS512) [53], and multi-scale CNN (MS-CNN) [19]. As can be seen from Table 4, the proposed SD-MS obtains the best mAP.…”

A Novel Multi-Model Decision Fusion Network for Object Detection in Remote Sensing Images

“…However, the aforementioned models can not be directly utilized for geospatial object detection, because the properties of remote sensing images and natural images are different and the direct application of those models to remote sensing images is not optimal. Researchers have done a lot of work in applying CNN-based models to detect geospatial objects in remote sensing images and achieved remarkable consequences [4,[15][16][17][18][19][20][21][22][23][24][25]45]. For example, the work in [4] utilized a hyperregion proposal network (HRPN) and a cascade of boosted classifiers to detect vehicles in remote sensing images.…”

“…Therefore, it is important for us to choose a method to extract features for object detection in remote sensing images. Currently, because of the advantage of directly generating more powerful feature representations from raw image pixels through neural networks, deep learning methods, especially CNN-based [4,[8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25], are recognized as predominate techniques for extracting features in object detection. Therefore, we select a CNN-based approach to extract features for object detection in optical remote sensing images.…”