CSCI 385 - Advances in Biometrics and Local AI

3 Credits
This course builds on the fundamentals of biometrics introduced in earlier coursework and connects them with advanced concepts from computer vision and pattern recognition. Students will learn biometric techniques currently used in industry and research, including feature extraction methods, biometric matching pipelines, and pattern recognition algorithms, while examining emerging trends and ongoing research in the field. In addition, the course introduces practical implementation of local artificial intelligence, including local large language models (LLMs), on on-premises computing environments (laptops, workstations, and other internal systems). Students will learn how to acquire, configure, and run modern open models locally, integrate local inference into software applications through programmatic interfaces, benchmark performance, and apply responsible data handling practices-particularly in privacy-sensitive biometric contexts.

Fees: Additional course fees apply.
Prerequisite(s): CSCI 380 and CSCI 410
Course Learning Outcomes:
1. Implement a trending biometric algorithm for a specific modality.
2. Critique the distinct biometric algorithms currently being used in industry and research.
3. Evaluate government applications of biometric authentication, including those used by the FBI, NIST, and ANSI.
4. Summarize current biometric research trends as well as individual researchers and research groups.
5. Configure local AI inference workflows (including local large language models) on on-premises computing environments using controlled test cases, output-validation criteria, and responsible data handling practices.
6. Develop a software application incorporating local biometric and/or LLM inference via programmatic interfaces (e.g., Python), including basic automation and performance benchmarking.