Search space reduction for person recognition using soft biometrics

A wide array of person recognition systems exists that perform automatic recognition based on different physiological and behavioral characteristics like face, fingerprint, iris, gait etc. Our work is based on implementation of a Soft Biometrics based recognition system. Soft Biometrics can be defined as traits that contain some information about the user, however, lack the necessary discrimination power to recognize a person effectively. Here, we analyze the effectiveness of a Soft Biometrics based recognition system. Thus, we demonstrate in our study how Soft Biometrics can be used in the task of person recognition by effectively reducing the search space in a database. We also demonstrate some of the limitations of this approach.