Missing data handling is an important preparation step for most data discrimination or mining tasks. Inappropriate treatment of missing data may cause large errors or false result...
Pattern detection problems require a separation between two classes, Target and Clutter, where the probability of the former is substantially smaller compared to that of the latte...
This paper investigates the feasibility of person identification based on feature points extracted from palmprint images. Our approach first extracts a set of feature points along...
This paper reports the application of a recently developed shape-from-shading technique to estimate facial pose. The shape-from-shading algorithm uses a new geometric technique fo...
Kwang Nam Choi, Philip L. Worthington, Edwin R. Ha...
As a new pattern classification method, nearest feature line (NFL) provides an effective way to tackle the sort of pattern recognition problems where only limited data are availab...
In this paper, general guidelines and specific algorithms for code transformations between breadth-first (BF) and depth-first (DF) linear quadtrees are proposed. Each algorithm ha...