Abstract

Roman imperial portraits are commonly recognized and identified by relating specific sculptures to commonly accepted types. The feature that has played a central role in this regard has been hairstyle. Based on the portrayed emperor’s coiffure and facial hair (or lack thereof), scholars have connected portraits to known classifications and prototypes. However, by focusing mainly on hairstyle, this method runs the risk of excluding portraits that do not adhere to the recognized repertoire as possible representations of Roman emperors altogether. Advances in digital art history and computational methods, however, may help solve this problem. By using DCNN-based methods, this article investigates the possibilities of digital identification of Roman imperial portraiture at a quantitative level, without using hairstyle. It demonstrates that there are sufficient visual markers in imperial portraits, aside from hairstyle, that allow facial recognition software to classify portraits when hairstyle is excluded.

DOI: https://doi.org/10.11588/dahj.2025.10.103033

AuthorS

Sam Heijnen

holds a PhD in Roman history (Radboud University, the Netherlands). His research interests include Roman history and Roman imperial portraiture. His titles include Living Up to Expectations: Hadrian’s Military Representation in Freestanding Sculpture, BABesch (2020), and Statues in Dialogue: Visual Similarities in “Grown” Roman Imperial Statue Groups from the Greek East, Ancient Society (2021).

Nova Hadi Lestriandoko

is a Senior researcher at the Research Center for Artificial Intelligence and Cyber Security (BRIN - National Research and Innovation Agency, Indonesia). He is also a PhD student at the Data Management and Biometrics Group (University of Twente, the Netherlands). His research interests include face recognition, image processing, and artificial intelligence. His titles include The Contribution of Different Face Parts to Deep Face Recognition, Frontiers in Computer Science (2022), and The Impact of Eyebrows Region on Deep Face Recognition, IC3INA (2023).

Sven Betjes

is a postdoctoral researcher at Radboud University (the Netherlands). His research focuses on the way ancient expressions of power were culturally embedded, which he especially studies for coinage and road-building in the Roman empire. His titles include Hadrian, the Pantheon and the Ideological Resonance of the Third Consulship, American Journal of Numismatics (2021), and Moneta, the memory of Roman coinage. The storage of dies (or coins) at the mint of Rome, The Journal of Archaeological Numismatics (2022).

Olivier Hekster

is Professor of Ancient History at Radboud University (the Netherlands). He studies the role of ideology in ancient Rome and the representation of Roman emperors in particular. The emphasis is on the various media that the emperors used to express their power (e.g. coins, inscriptions, portraits and buildings) and how the population in the Roman Empire responded to this. His titles include Emperors and Ancestors. Roman Rulers and the Constraints of Tradition, Oxford University Press (2015), and Caesar Rules. The Emperor in the Changing Roman World (c. 50 BC- AD 565), Cambridge University Press (2023).

Luuk Spreeuwers

is Associate Professor at the Data Management and Biometrics Group of the Department of EEMCS of the University of Twente (the Netherlands). His expertise involves digital image processing and analysis, medical image analysis, biometrics, and pattern recognition in general. His titles include Breaking the 99% Barrier: Optimisation of 3D Face Recognition, IET biometrics (2015), Practical Evaluation of Face Morphing Attack Detection Methods, in: Handbook of Digital Face Manipulation and Detection, Springer (2022).

Florens de Wit

is a Researcher Biometrics at the Data Management and Biometrics Group of the Department of EEMCS of the University of Twente (the Netherlands). His main area of expertise is performance evaluation of biometric systems. His titles include Biometric Testing: Aligning Standards and Practice. Paper presented at 42nd WIC Symposium on Information Theory and Signal Processing in the Benelux (2022).