Linear Gaussian Bounding Box Representation and Ring-Shaped Rotated Convolution for Oriented Object Detection
This work addresses a specific problem in computer vision for oriented object detection, offering incremental improvements over existing methods.
The paper tackles the boundary discontinuity problem in oriented object detection by proposing a linear Gaussian bounding box representation (LGBB) that avoids this issue and improves numerical stability, and introduces ring-shaped rotated convolution (RRC) to enhance feature aggregation, achieving state-of-the-art performance with effective integration into various models.
In oriented object detection, current representations of oriented bounding boxes (OBBs) often suffer from boundary discontinuity problem. Methods of designing continuous regression losses do not essentially solve this problem. Although Gaussian bounding box (GBB) representation avoids this problem, directly regressing GBB is susceptible to numerical instability. We propose linear GBB (LGBB), a novel OBB representation. By linearly transforming the elements of GBB, LGBB avoids the boundary discontinuity problem and has high numerical stability. In addition, existing convolution-based rotation-sensitive feature extraction methods only have local receptive fields, resulting in slow feature aggregation. We propose ring-shaped rotated convolution (RRC), which adaptively rotates feature maps to arbitrary orientations to extract rotation-sensitive features under a ring-shaped receptive field, rapidly aggregating features and contextual information. Experimental results demonstrate that LGBB and RRC achieve state-of-the-art performance. Furthermore, integrating LGBB and RRC into various models effectively improves detection accuracy.