Abstract: Current deep learning-based image matching methods generally face a trade-off between computational efficiency and matching accuracy: although existing approaches have significantly improved ...
Fine-grained classification is particularly challenging due to low inter-class variability. While many approaches have been proposed to classify fine-grained images, some overlook the correlation ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results