Rotation Equivariant Feature Image Pyramid Network for Object Detection in Optical Remote Sensing Imagery

TL;DR

This paper introduces REFIPN, a rotation equivariant feature pyramid network that improves object detection in aerial images by effectively handling scale and orientation variations, especially for small objects.

Contribution

The paper proposes a novel rotation equivariant convolution-based image pyramid network that enhances multi-scale and multi-orientation feature extraction for aerial object detection.

Findings

Achieves state-of-the-art performance on aerial benchmarks.

Improves detection accuracy for small-scale objects.

Maintains efficiency while handling diverse object orientations.

Abstract

Detection of objects is extremely important in various aerial vision-based applications. Over the last few years, the methods based on convolution neural networks have made substantial progress. However, because of the large variety of object scales, densities, and arbitrary orientations, the current detectors struggle with the extraction of semantically strong features for small-scale objects by a predefined convolution kernel. To address this problem, we propose the rotation equivariant feature image pyramid network (REFIPN), an image pyramid network based on rotation equivariance convolution. The proposed model adopts single-shot detector in parallel with a lightweight image pyramid module to extract representative features and generate regions of interest in an optimization approach. The proposed network extracts feature in a wide range of scales and orientations by using novel convolution filters. These features are used to generate vector fields and determine the weight and angle of the highest-scoring orientation for all spatial locations on an image. By this approach, the performance for small-sized object detection is enhanced without sacrificing the performance for large-sized object detection. The performance of the proposed model is validated on two commonly used aerial benchmarks and the results show our proposed model can achieve state-of-the-art performance with satisfactory efficiency.