116 datasets found

The VoxCeleb dataset is a large-scale speaker identification dataset, used to evaluate the performance of face recognition systems.

FaceNet is a deep neural network for face recognition. It uses a convolutional neural network (CNN) architecture and is trained on a large dataset of faces.

Blind face restoration dataset

Face recognition and person re-identification using paired image-attribute data, where the attributes (i.e., soft biometrics) are only available during the training phase.

Sum-Product Networks (SPNs) are hierarchical, graphical models that combine benefits of deep learning and probabilistic modeling.

The JAFFE dataset is a collection of images of Japanese faces, each with a specific facial expression. The dataset was created to test the psychological plausibility of a...

This dataset is used to train and evaluate the DCFace model.

This dataset is used to estimate the biometric capacity of generative face models.

Labeled Faces in the Wild: A database for studying face recognition in unconstrained environments.

A light-weight low-resolution face gender classification method, called FaceHop, is proposed in this research.

The dataset is used for evaluating the fairness of people annotations.

The dataset is used for evaluating the dual-glance model for deciphering social relationships.

The dataset used for training and testing the proposed method for restoring a single face image distorted by atmospheric turbulence.

FairFace is a dataset of face images for fairness audit and evaluation, intentionally constructed to be approximately racially balanced.

The LAION-Face dataset consists of 50 million image-text pairs to ensure diversity.

The Specs on Faces (SoF) dataset is a challenging dataset for face detection, recognition, and classification.

The Face Recognition Technology (FERET) dataset consists of 14,126 images for 1199 different persons.

A dataset for image content changes using deep feature interpolation.

A dataset for facial landmark tracking in-the-wild.

The FG-NET1 dataset contains 1002 images colored resolution, quality, or gray with variation in illumination, viewpoint, and expression. There are 82 different subjects with...