Genetic search for face detection and verification

We investigate the application of genetic algorithms (GAs) to search for the face of a particular individual in a two-dimensional intensity image. This problem has many potential applications such as locating a person in a crowd from pictures taken by surveillance cameras. There are two steps in solving this problem: first, faces regions must be extracted from the image (face detection) and second, candidate faces must be compared against the race of interest (face verification). Without any a priori knowledge about location and size of a face in an image, every possible image location and face size must be considered, leading to a very large search space. In addition, face detection or verification invariant to lighting conditions, facial expression and pose make the search space even larger and more complex. In the paper, we propose using GAs to search the image efficiently. Specifically, we use GAs to find image sub-windows that contain the face of interest. Each sub-window is evaluated using a fitness function which contains two terms: the first term favors sub-windows containing faces while the second term favors sub-windows containing faces similar to the face of interest. Both terms have been defined using ideas from the method of "eigenfaces". A set of increasingly complex scenes demonstrate the performance of the genetic search approach.