In this study, an improved end-to-end framework for instance segmentation of military camouflaged targets, referred to as MilInst, is proposed. The builds upon SparseInst method developed by Cheng et al . [27]. Several improvements are introduced enhance the model’s performance. First, Receptiv...

Abstract Recognizing 3D part instances from a point cloud is crucial for structure and scene understanding. Several learning-based approaches use semantic segmentation instance center prediction as training tasks fail to further exploit the inherent relationship between shape semantics instances. In this paper, we present new method segmentation. Our exploits fuse nonlocal features, such predic...

Instance recognition is rapidly advanced along with the developments of various deep convolutional neural networks. Compared to architectures networks, training process, which also crucial success detectors, has received relatively less attention. In this work, we carefully revisit standard practice and find that detection performance often limited by imbalance during generally consists in thre...

Long-tailed class distributions are prevalent among the practical applications of object detection and instance segmentation. Prior work in long-tail segmentation addresses imbalance losses between rare frequent categories by reducing penalty for a model incorrectly predicting label. We demonstrate that heavily suppressed correct background predictions, which reduces probability all foreground ...

Existing instance segmentation models mostly pay less attention to the targeted characteristics of ships in synthetic aperture radar (SAR) images, which hinders further accuracy improvements, leading poor performance more complex SAR image scenes. To solve this problem, we propose a hybrid task cascade plus (HTC+) for better ship segmentation. Aiming at specific task, seven techniques are propo...