Project
Forensic ico-biometrics (PhD Position)
The purpose of this research is to develop fully automated, robust and accurate computational methods for morpho-phenotyping of human anatomy on large-scale two- and three-dimensional in-vivo medical imaging (iconic) data of the craniofacial skeleton. The main application is direct biological profiling (biometrics), which consists of estimating sex, age, stature, and ancestry. Through direct intra- and extra-muros collaborations with forensic odontology and forensic anthropology departments, we have access to extensive image databases of imaging data (Orthopantographic (OPG) X-rays and (Cone-Beam) Computed Tomography (CB)CT) for age and ancestry estimation. The main methodological novelty is to useend-to-end Deep Learning to substitute for the engineered, often ad-hoc, approaches to segmentation, landmarking and warping.