We introduce a new method that characterizes typical local image features (e.g., SIFT [9], phase feature [3]) in terms of their distinctiveness, detectability, and robustness to i...
For the improvement of iris recognition system performance, the filtering algorithm that picks out counterfeit and noisy data is very important. In this paper, as a part of preproc...
Abstract. In this paper, we describe an approach to activity recognition, which is based on a self-organizing, ad hoc network of body-worn sensors. It makes best use of the availab...
In this paper, we propose a human action recognition system suitable for embedded computer vision applications in security systems, human-computer interaction and intelligent envi...
Many existing biometric systems collect ancillary information like gender, age, height, and eye color from the users during enrollment. However, only the primary biometric identiï¬...