/RealFaceDataset

The Real Face Dataset is a pedestrian face detection benchmark dataset in the wild, containing over 11000 images and over 55000 detected faces in various ambient conditions.

RealFaceDataset

cover

Introduction

This repository contains a data collection of over 11000 pedestrian images taken in the wild. The data was collected from a variety of sources in the hope of creating a benchmark dataset for pedestrian face detection / recognition research. The dataset contains over 55000 detected faces. The dataset is freely available for academic research purposes.

Race Composition

Race % of total faces
Asian 13.43%
Black or African American 5.22%
Hispanic or Latino 5.97%
White 75.37%

Download

Download the Real Face sataset through Google Drive, unzip the downloaded file and you will get the following structure:

.
├── 1.jpg                   # Image file
├── 1.xml                   # Annotations file
├── ...
├── 11437.jpg
└── 11437.xml

Annotation format

<annotation>
	<object>
		<name>face</name>
		<pose>Unspecified</pose>
		<truncated>0</truncated>
		<difficult>0</difficult>
		<bndbox>
			<xmin>773</xmin>
			<ymin>437</ymin>
			<xmax>796</xmax>
			<ymax>468</ymax>
		</bndbox>
	</object>
	...
</annotation>

Contact

For questions and result submission, please contact Leonardo Ramos at 08-10921@usb.ve