five

zblasingame/DiM-FRLL-Morphs

收藏
Hugging Face2024-12-17 更新2024-12-21 收录
下载链接:
https://hf-mirror.com/datasets/zblasingame/DiM-FRLL-Morphs
下载链接
链接失效反馈
官方服务:
资源简介:
DiM FRLL Morphs数据集是通过DiM、Fast-DiM和Greedy-DiM算法在FRLL数据集上生成的变形图像集合。每个变形算法生成了500个独特的变形图像。数据集的组织结构为`morphs/<morphing_algorithm>/<id_a>_<id_b>.png`,其中`<morphing_algorithm>`是用于创建变形的特定算法,`<id_a>`和`<id_b>`是用于构建变形的FRLL数据集中真实图像的ID标签。图像分辨率为256x256,并且使用了dlib进行对齐和裁剪。数据集的使用注意事项包括在评估变形性能时不要与原始真实图像进行比较。

The DiM FRLL Morphs dataset is a collection of morphed images generated via the DiM, Fast-DiM, and Greedy-DiM algorithms on the FRLL dataset. Each morphing algorithm produced 500 unique morphs. The dataset is organized as `morphs/<morphing_algorithm>/<id_a>_<id_b>.png`, where `<morphing_algorithm>` is the specific algorithm used to create the morphs, and `<id_a>` and `<id_b>` are the ID labels for the bona fide images from FRLL used to construct the morphs. The images have a resolution of 256x256 and were aligned and cropped using dlib. Important usage notes include not comparing the morphs against the original bona fide images when evaluating morphing performance.
提供机构:
zblasingame
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作