Caltech 10k Web Faces

Caltech 10k Web Faces数据集是由从Google搜索获得的图片组成的数据集。每个图片中眼睛、鼻子和嘴巴的中心坐标在相关文件中提供。相关信息可用于对齐和裁剪人脸、人脸识别算法的基准，数据集共有10,524幅不同分辨率的图像，并用于不同领域。
这个数据集由加州理工2005年发布，相关论文是《剪枝训练集以学习对象类别》。

The Caltech 10k Web Faces dataset is composed of images acquired via Google searches. Central coordinates of the eyes, nose, and mouth for each individual image are provided in the accompanying files. This relevant information can be utilized for face alignment and cropping, as well as serving as a benchmark for face recognition algorithms. The dataset contains a total of 10,524 images with varying resolutions and has been applied across multiple research fields. It was released by the California Institute of Technology in 2005, and its associated paper is titled *Pruning the Training Set for Learning Object Categories*.

搜集汇总

背景与挑战

背景概述

Caltech 10k Web Faces数据集是由加州理工学院2005年发布的，包含10,524幅从Google搜索获得的图片，每张图片标注了眼睛、鼻子和嘴巴的中心坐标，适用于人脸对齐、裁剪和人脸识别算法的基准测试。