Strip R-CNN: Large Strip Convolution for Remote Sensing Object Detection

TL;DR

This paper introduces Strip R-CNN, a novel remote sensing object detection network utilizing large orthogonal strip convolutions to effectively detect objects with various aspect ratios, achieving state-of-the-art results.

Contribution

The paper proposes a new network architecture, Strip R-CNN, that leverages large strip convolutions for improved feature extraction and localization in remote sensing object detection.

Findings

Achieves 82.75% mAP on DOTA-v1.0, setting a new state-of-the-art.

Demonstrates superior performance on multiple benchmarks like FAIR1M, HRSC2016, and DIOR.

Shows that large strip convolutions effectively capture spatial information for high aspect ratio objects.

Abstract

While witnessed with rapid development, remote sensing object detection remains challenging for detecting high aspect ratio objects. This paper shows that large strip convolutions are good feature representation learners for remote sensing object detection and can detect objects of various aspect ratios well. Based on large strip convolutions, we build a new network architecture called Strip R-CNN, which is simple, efficient, and powerful. Unlike recent remote sensing object detectors that leverage large-kernel convolutions with square shapes, our Strip R-CNN takes advantage of sequential orthogonal large strip convolutions in our backbone network StripNet to capture spatial information. In addition, we improve the localization capability of remote-sensing object detectors by decoupling the detection heads and equipping the localization branch with strip convolutions in our strip head. Extensive experiments on several benchmarks, for example DOTA, FAIR1M, HRSC2016, and DIOR, show that our Strip R-CNN can greatly improve previous work. In particular, our 30M model achieves 82.75% mAP on DOTA-v1.0, setting a new state-of-the-art record. Our code will be made publicly available.Code is available at https://github.com/YXB-NKU/Strip-R-CNN.