Dual Appearance-Aware Enhancement for Oriented Object Detection

• Published in: IEEE transactions on geoscience and remote sensing Vol. 62; pp. 1 - 14
• Main Authors: Gong, Maoguo, Zhao, Hongyu, Wu, Yue, Tang, Zedong, Feng, Kai-Yuan, Sheng, Kai
• Format: Journal Article
• Language: English
• Published: New York IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Anchor-free; appearance sensitivity detector (ASDet); Aspect ratio; Detection; Detectors; dual appearance-aware (DA) loss; Feature extraction; Localization; Object detection; Object recognition; oriented object detection; Remote sensing; remote-sensing images; Sensitivity; Sensors; Task analysis; Training; Vectors; weak feature extraction
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2023.3344195

Oriented detectors have become the mainstream of object detection in remote-sensing images since they provide more precise bounding boxes and contain less background. However, there remain several challenges that restrict the detection performance and need to be tackled. This article focuses on the following two aspects: 1) numerous tiny objects in remote-sensing images pose a challenge for the detectors pursuing high recall and accurate localization and 2) specific categories with large aspect ratios and arbitrary angles also trouble the regression of the detectors. We attempt to alleviate the above problems by constructing a weak feature extraction network (WFEN) and a dual appearance-aware (DA) loss. Specifically, WFEN is used to extract hierarchical weight vectors for multiscale feature layers by employing a lightweight convolutional module, aiming to fuse activation features distributed in different layers and provide pure features for subsequent regression and classification. DA loss is tailored to regressions of tiny and slender objects by dynamically modulating the associated loss on objects with various appearances, which consists of two auxiliary losses, termed scale-aware loss <inline-formula> <tex-math notation="LaTeX">{\mathcal {L}}_{S} </tex-math></inline-formula> and aspect-ratio-aware loss <inline-formula> <tex-math notation="LaTeX">{\mathcal {L}}_{A} </tex-math></inline-formula>. These two components can contribute to each other, that is, the former provides more accurate features for detection tasks, while the latter can reciprocate the former by imposing constraints on crucial objects, and together constitute an appearance sensitivity detector (ASDet). Extensive experiments on three public datasets demonstrate that our ASDet outperforms all refine-stage detectors in terms of accuracy while maintaining the superior inference speed of single-stage counterparts.