4.7 Article

Deep Saliency Smoothing Hashing for Drone Image Retrieval

出版社

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TGRS.2023.3255302

关键词

Drones; Codes; Feature extraction; Image retrieval; Smoothing methods; Signal processing algorithms; Remote sensing; Deep hashing; drone image retrieval; local fine-grained features; saliency information

向作者/读者索取更多资源

This article proposes a novel deep saliency smoothing hashing (DSSH) algorithm to learn effective hash codes for drone image retrieval by leveraging saliency capture mechanism, distribution smoothing term, global features, and local fine-grained features. Extensive experiments demonstrate that the DSSH algorithm can further improve retrieval performance compared with other deep hashing algorithms.
Deep hashing algorithms are widely exploited in retrieval tasks due to their low storage and retrieval efficiency. Most of them focus on global feature learning, while neglecting local fine-grained features and saliency information for drone images. In this article, we tackle these dilemmas with a novel deep saliency smoothing hashing (DSSH) algorithm, which can leverage saliency capture mechanism, distribution smoothing term, global features, and local fine-grained features to learn effective hash codes for drone image retrieval. The DSSH algorithm first designs an information extraction module to capture global features and local fine-grained features for drone images. Meanwhile, a saliency capture module is proposed to perform information interaction attention and visual enhancement attention, which can capture the saliency area of drone images effectively. On top of the two paths, a novel objective function is designed to preserve the similarity of hash codes, smooth the distribution of drone image datasets, and reduce the quantization errors between hash codes and hash-like codes concurrently. Extensive experiments on the Drone Action Dataset and ERA Drone Dataset demonstrate that the DSSH algorithm can further improve the retrieval performance compared with other deep hashing algorithms.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据