Abstract:Deep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresolution remote sensing images, where the quality is defined by the Intersection over Union (IoU) threshold used in training. In addition, the imbalance between the positive and negative samples across multiresolution images worsens the detection precision. Recently, it was found that a Cascade region-based convolutional neural network (R-CNN) can potentially achieve a higher quality of detection by introducing a cascaded three-stage structure using progressively improved IoU thresholds. However, the performance of Cascade R-CNN degraded when the fourth stage was added. We investigated the cause and found that the mismatch between the ROI features and the classifier could be responsible for the degradation of performance. Herein, we propose a Cascade R-CNN++ structure to address this issue and extend the three-stage architecture to multiple stages for general use. Specifically, for cascaded classification, we propose a new ensemble strategy for the classifier and region of interest (RoI) features to improve classification accuracy at inference. In localization, we modified the loss function of the bounding box regressor to obtain higher sensitivity around zero. Experiments on the DOTA dataset demonstrated that Cascade R-CNN++ outperforms Cascade R-CNN in terms of precision and detection quality. We conducted further analysis on multiresolution remote sensing images to verify model transferability across different object scales.

CRRCNN: Cascade Rotational RCNN for Dense Arbitrary-Oriented Object Detection

Direction-Sensitivity Features Ensemble Network for Rotation-Invariant Face Detection.

RINet: Efficient 3D Lidar-Based Place Recognition Using Rotation Invariant Neural Network

Rotation-Invariant Face Detection with Multi-task Progressive Calibration Networks.

R2CNN: Rotational Region CNN for Orientation Robust Scene Text Detection

Rotated Cascade R-CNN: A Shape Robust Detector with Coordinate Regression

R2CNN++: Multi-Dimensional Attention Based Rotation Invariant Detector with Robust Anchor Strategy.

Improved Oriented Object Detection in Remote Sensing Images Based on a Three-Point Regression Method

Structurally re-parameterized rotation detector for arbitrary-oriented objects in high-resolution remote sensing images

Point RCNN: An Angle-Free Framework for Rotated Object Detection

Adaptive Rotated Convolution for Rotated Object Detection

Oriented R-CNN for Object Detection

Oriented R-CNN and Beyond

Effective Rotate: Learning Rotation-Robust Prototype for Aerial Object Detection

Arbitrary-angle bounding box based location for object detection in remote sensing image

Parallel Cascade R-CNN for Object Detection in Remote Sensing Imagery

RiDOP: A Rotation-Invariant Detector with Simple Oriented Proposals in Remote Sensing Images

Rifd-Cnn: Rotation-Invariant And Fisher Discriminative Convolutional Neural Networks For Object Detection

RotaTR: Detection Transformer for Dense and Rotated Object

High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors.

C-RPNs: Promoting object detection in real world via a cascade structure of Region Proposal Networks.