Glint360K

该数据集名为Glint360K，被用作蒙面人脸识别任务的基础训练集。此外，该数据集的应用旨在实现蒙面人脸识别和标准人脸识别任务之间的平衡性能。所涉及的任务是蒙面人脸识别。

Glint360K is a dataset employed as the foundational training set for masked face recognition tasks. Moreover, it is developed to achieve balanced performance between masked face recognition and standard face recognition tasks. The core task addressed by this dataset is masked face recognition.

搜集汇总

背景与挑战

背景概述

Glint360K是一个大规模人脸识别数据集，包含17091657张图像和360232个个体，是目前最大且最干净的人脸识别数据集之一。该数据集支持非商业研究用途，并提供了多种下载方式和预训练模型。