Abstract: A key issue of existing deep-learning-based object detection methods in remote sensing images is that they often struggle to differentiate the background and small object regions due to ...
Abstract: This paper improves upon the Pix2Seq object detector by extending it for videos. In the process, it introduces a new way to perform end-to-end video object detection that improves upon ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results