FRED：迈向航空图像目标检测中的完全旋转等变性
FRED: Towards a Full Rotation-Equivariance in Aerial Image Object Detection
AI总结：
本文提出完全旋转等变的定向目标检测器FRED，通过解耦分类与定位任务、用旋转等变向量表示边界框并作为可变形卷积偏移量，在DOTA数据集上以16%参数量取得可比或更优性能。
摘要：
旋转等变性是定向目标检测中一个至关重要但极具挑战性的性质。通用目标检测器由于传统卷积神经网络的平移等变性，天然能够利用对空间平移的鲁棒性，然而实现旋转等变性仍然是一个难以达成的目标。当前检测器采用各种对齐技术来获得旋转不变特征，但仍依赖高容量模型以及对所有可能旋转进行的大量数据增强。在本文中，我们提出了一种完全旋转等变的定向目标检测器（FRED），其从图像到边界框预测的整个过程都严格满足等变性。具体而言，我们将不变任务（目标分类）与等变任务（目标定位）解耦，以实现端到端的等变性。我们将边界框表示为一组旋转等变向量，从而实现旋转等变定位。此外，我们利用这些旋转等变向量作为可变形卷积中的偏移量，从而增强了现有空间自适应能力的优势。凭借完全的旋转等变性，我们的FRED相比现有方法对图像级旋转表现出更高的鲁棒性。进一步地，我们通过实验表明，FRED向非轴对齐学习又迈进了一步。与最先进方法相比，我们提出的方法在DOTA-v1.0上取得了可比的性能，并在DOTA-v1.5上以1.5 mAP的优势超越现有方法，同时将模型参数量显著降低至16%。
英文摘要：
Rotation-equivariance is an essential yet challenging property in oriented object detection. While general object detectors naturally leverage robustness to spatial shifts due to the translation-equivariance of the conventional CNNs, achieving rotation-equivariance remains an elusive goal. Current detectors deploy various alignment techniques to derive rotation-invariant features, but still rely on high capacity models and heavy data augmentation with all possible rotations. In this paper, we introduce a Fully Rotation-Equivariant Oriented Object Detector (FRED), whose entire process from the image to the bounding box prediction is strictly equivariant. Specifically, we decouple the invariant task (object classification) and the equivariant task (object localization) to achieve end-to-end equivariance. We represent the bounding box as a set of rotation-equivariant vectors to implement rotation-equivariant localization. Moreover, we utilized these rotation-equivariant vectors as offsets in the deformable convolution, thereby enhancing the existing advantages of spatial adaptation. Leveraging full rotation-equivariance, our FRED demonstrates higher robustness to image-level rotation compared to existing methods. Furthermore, we show that FRED is one step closer to non-axis aligned learning through our experiments. Compared to state-of-the-art methods, our proposed method delivers comparable performance on DOTA-v1.0 and outperforms by 1.5 mAP on DOTA-v1.5, all while significantly reducing the model parameters to 16%.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
