• IdentAI Database

    The data acquisition started in October 2024 in Granada, Spain, during the 20th Meeting of the International Association for Craniofacial Identification (IACI) where it was first introduced. This conference was held at the Health Sciences Campus belonging to the University of Granada, and at this location we captured images of 26 individuals. The setup was then moved to the Andalusian Interuniversity Institute in Data Science and Computational Intelligence (DASci) building, also in Granada, where images of other 20 individuals were taken. Finally in April 2025 the setup was moved to Ponferrada, Spain, where an office of Panacea is located. At this place we have gathered facial images from 65 individuals up until April 2026. 

    THE DATASET

    Acquiring facial images in the three modalities – standardized photos, recordings, and 3D models, – is still ongoing. Up to now, April 2026, we have been able to gather images from a total of 111 individuals, 51 female and 60 male participants. The images have been captured with varied lighting conditions, and some of them have both outdoor-indoor scenarios, as well as varied degrees of occlusion including sunglasses, eyepatch, a cap, and a facemask.

    This dataset includes 644 3D facial models showing neutral and varied expressions (smile without teeth, smile with teeth, upset, angry, and surprised). Additionally, the collection contains 231 standardized facial photographs and approximately 11.4 hours of video recordings, totaling around 920,000 frames.

    EXAMPLES

    Figure 1. Frames extracted from the CCTV recordings different days, locations, occluding objects and lighting conditions.
    Figure 2. Standardized photographs, one frontal and two profile images (L/R)
    Figure 3. 3D high-resolution facial models depicting three facial expressions (surprise, neutral, happy without teeth)

    DATA SAMPLE

    If you are interested in accessing a portion of our dataset for testing, you can request it by filling out this form.