CSC 577 : Pattern Recognition

This course introduces pattern recognition methods and theory discussing topics such as feature extraction, statistical classification, neural networks, fuzzy logic, support vectors, linear discriminant analysis, principal component analysis, clustering, and unsupervised learning.
Students implement algorithms, apply methods to selected problems, and document findings.

 

Student Work :

 

Students completed several projects where they built a variety of pattern classifiers that could be used on manufactured and also on real-world data. Python, SciPy, and NumPy along with MatPlotLib were heavily used as open-source scientific programming tools for coding, implementing, and testing classifiers. Many techniques were tested on actual face data using Active Appearance Model parameters generated from an ongoing face-aging-related project here at UNCW. Students built gender and ethnicity classifiers using this real-world data based on over 500 faces. Students also completed final projects on a variety of topics, investigating topics and implementing classifiers that they had designed including image searches from Flickr, audio analysis of dub-step and drum-bass music, and others.