UMDFaces: An Annotated Face Dataset for Training Deep Networks

TL;DR

The paper introduces UMDFaces, a large publicly available face dataset with extensive annotations, and proposes a new face recognition evaluation protocol to facilitate progress in the field.

Contribution

It provides a new large-scale face dataset with detailed annotations and a novel evaluation protocol for face recognition research.

Findings

UMDFaces contains 367,888 faces of 8,277 subjects.

The dataset includes pose, keypoints, gender annotations, and human-verified labels.

Comparison shows UMDFaces quality exceeds similar public datasets.

Abstract

Recent progress in face detection (including keypoint detection), and recognition is mainly being driven by (i) deeper convolutional neural network architectures, and (ii) larger datasets. However, most of the large datasets are maintained by private companies and are not publicly available. The academic computer vision community needs larger and more varied datasets to make further progress. In this paper we introduce a new face dataset, called UMDFaces, which has 367,888 annotated faces of 8,277 subjects. We also introduce a new face recognition evaluation protocol which will help advance the state-of-the-art in this area. We discuss how a large dataset can be collected and annotated using human annotators and deep networks. We provide human curated bounding boxes for faces. We also provide estimated pose (roll, pitch and yaw), locations of twenty-one key-points and gender information generated by a pre-trained neural network. In addition, the quality of keypoint annotations has been verified by humans for about 115,000 images. Finally, we compare the quality of the dataset with other publicly available face datasets at similar scales.