상명대 생물정보학 연구실

Development of quantitative evaluation system of forensic biometrics by using artificial intelligence

This research aims to establish a machine learning-based standard system for the quantitative evaluation of non-human DNA and biological markers used in forensic investigations.

- Machine Learning-Based Species Identification System for Forensic DNA

This study focuses on developing an automated species identification system for forensic investigations using non-human biological DNA and statistically verifying its validity. A likelihood-based quantification system is established and optimized for practical application.

Additionally, DNA sequences from various organisms, including zooplankton, microorganisms, and other flora and fauna found in domestic freshwater environmental samples, are identified and compiled into a dataset. This dataset serves as raw input for the species identification system.

- Quantitative Analysis System for Biometrics-Based Image Information

This research aims to develop a biometric analysis system that applies optimized statistical techniques from a forensic science perspective to image-based biometric data, such as handwriting and facial recognition.

The system will integrate handwriting authentication, identity recognition via facial features, and psychological assessment through facial analysis. Furthermore, it will incorporate quantitative likelihood ratio calculations to ensure forensic admissibility and evidentiary strength in legal proceedings.