UKBench, COREL, UCID
收藏资源简介:
本研究使用了三个数据集:UKBench、COREL和UCID。UKBench数据集包含1000张图像,分为250组,每组四张图像,用于评估隐私保护的图像检索方案。COREL和UCID数据集用于生成代码书,这些数据集独立于上传的加密图像,有助于提高检索准确性。整个数据集创建过程涉及图像分割、特征提取和代码书生成,旨在解决云环境中图像隐私保护和高效检索的问题。
This study employs three datasets: UKBench, COREL, and UCID. The UKBench dataset consists of 1000 images divided into 250 groups, with four images per group, and is used to evaluate privacy-preserving image retrieval schemes. The COREL and UCID datasets are utilized to generate codebooks; these datasets are independent of the uploaded encrypted images and help improve retrieval accuracy. The entire dataset creation process involves image segmentation, feature extraction, and codebook generation, aiming to address the issues of image privacy protection and efficient retrieval in cloud environments.

- 1A Privacy-Preserving Image Retrieval Scheme Using A Codebook Generated From Independent Plain-Image Dataset东京都立大学 · 2021年



