Approved for: 1 GIAC CPE
The identification of an unknown individual poses a difficult problem to investigators. DNA fingerprinting allows for an investigator to determine a person's identity based on their CODIS DNA profile. However, when that profile does not exist in a database, DNA still provides the ability to predict physical characteristics of the individual. Advanced DNA analytical methods have allowed investigations of genomic correlations to an individual's biographic ancestry, visible traits, age and more recently, an individual's craniofacial morphology. Ultimately, the potential exists for this type of technology to cross biometric domains. By attempting to predict the face of an unknown individual, this capability proposes the question of if and how this technology would benefit the face biometric community. This session will initiate the dialogue for the identity community to understand the technology behind the predicted appearance, its potential usefulness to the identity mission and the limitations of the current capabilities.
11:00-11:10
DNA Phenotyping and Kinship Determination
Ellen McRae
11:10-11:20
Matching DNA and Predictive Modeling of the Human Face
Mark Shriver
11:20-11:30
Predictive Association for Facial Morphology and Age Estimation
Nani Grimmer
11:30-11:40
DNA Prediction of Facial Attributes with Deep Learning
Matt Peterson
11:40-11:50
DNA2Face
Chris Boehnen
11:50-12:15
Panel Q&A