Fusion analysis of soft biometrics for recognition at a distance
Bingchen H. Guo, Mark S. Nixon, John N. Carter · 2018
There are many new soft biometrie approaches though few are used for identification, especially at a distance. We create a new soft biometric database containing the human face, body and clothing attributes at three different distances. One aim of this paper is to explore the potency of face, body and clothing for human recognition using fusion. Another aim is to explore the suitability for each modality at a distance: intuitively face is suited to near field identification and body when the subject is further away. The new database and the soft biometric features used are clarified and the influence of distance on soft biometric traits is investigated. Results show the individual advantages and disadvantages of face, body and clothing traits and demonstrate the superiority of fusion method in human recognition compared with single-modality.