Race and ethnicity labels in Morph II are self-reported, which is good practice—but they are coarse (only seven categories). A person identifying as "Black" could have vastly different facial features based on Afro-Caribbean, African American, or recent African immigrant backgrounds. This reduces the granularity of fairness analyses.
The MORPH II dataset has several applications: morph ii dataset
The Morph II dataset (often stylized as MORPH-II) is a large-scale, longitudinal dataset of facial images primarily designed for research on age progression and face recognition across time. Unlike static datasets that capture a single image per subject, Morph II contains multiple images of the same individuals taken over periods ranging from months to several years. Race and ethnicity labels in Morph II are
Created by Karl Ricanek Jr. and his team at the University of North Carolina Wilmington (UNCW), Morph II was released as an extension of the original MORPH dataset (Morph I). While the first version focused on a smaller, more constrained sample, Morph II exploded in scale and diversity, becoming one of the most cited resources in age-invariant face recognition. As of 2024, the dataset is not available
The MORPH II dataset is a comprehensive benchmark for evaluating face recognition systems and face morphing attacks. The dataset provides a diverse and challenging set of images, which can be used to evaluate the performance of face recognition systems and detect morphed images. The dataset has several applications in biometric security, face recognition, and face morphing attacks. However, it also presents several challenges and limitations, which must be carefully considered when using the dataset.
Unlike many modern face datasets that are freely downloadable, Morph II is restricted. Researchers must submit a formal request to the original authors (via the UNCW face aging lab), sign a usage agreement, and often pay a nominal fee to cover distribution costs. The restrictions exist for two reasons:
As of 2024, the dataset is not available on common repositories like Kaggle or Hugging Face. However, many papers that cite Morph II provide "Morph-II-like" subsets or synthetic derivatives to enable reproducibility without redistributing the original data.