Abstract: The remote sensing image object detection has advanced significantly; yet, small object detection remains challenging due to their limited size and varying scales. Furthermore, real-world ...
Abstract: Traditional object detection models often lose the detailed outline information of the object. To address this problem, we propose the Fourier Series Object Detection (FSD). It encodes the ...
The DINOv3 codebase includes several task-specific heads for a variety of tasks, including classification, depth estimation, object detection, and semantic segmentation. The initial focus of this ...
The inherent noisy and sparse characteristics of radar data pose challenges in finding effective representations for 3D object detection. In this paper, we propose RadarDistill, a novel knowledge ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback