Skip to header Skip to Content Skip to Footer
UNCW professor Dr. Karl Ricanek (right) works with his team of research associates Ben Barbour and Amrutha Sethuram developing and creating software to help with facial recognition for government bureaus such as the FBI.
img_30

MORPH

The MORPH Longitudinal Database is the largest longitudinal facial recognition database in the world. Available in a variety of sizes and formats to fit any use case, the longitudinal database has images of people of different races and genders over time, complete with extensive metadata. MORPH datasets are commonly used for AI and machine learning training, facial recognition research, and a wide range of other technological applications. Explore our products below. 

MORPH Products

MORPH 5

The MORPH 5 dataset is our premier longitudinal dataset, comprising over 521,000 images from over 106,000 subjects. MORPH’s diverse range of ancestry includes subjects of African, Asian, European, Hispanic, Indian, and Caucasian descent, spanning from ages 18-89. The dataset can help overcome algorithmic bias by its diversity of subjects and inclusion of robust data, such as height and weight. The dataset contains metadata for age, gender, race, height, weight, and eye coordinates.

MORPH 4

The MORPH 4 dataset shares the same benefits as MORPH 5 but with fewer images (402,000) and fewer subjects (67,600), creating a robust yet cost-effective solution.

MORPH Multimedia

MORPH MultiMedia is a de novo dataset that includes high-res images from multiple angles, videos, and mobile captures, providing a comprehensive view of facial expressions of 3,000+ subjects. The dataset spans a wide range of racial and ethnic groups, biological sexes, ages and extensive metadata.

MORPH MultiMediaLite

MORPH MultiMediaLite contains the same images as MORPH MultiMedia, but without the video and mobile captures, creating a more cost-effective yet robust dataset.