A novel approach is proposed to analyzing and tracking the motion of structured deformable shapes, which consist of multiple correlated deformable subparts. Since this problem is ...
Approaches analyzing local characteristics of an image prevail in image restoration. However, they are less effective in cases of restoring images degraded by large size point spr...
As biometric authentication systems become more prevalent, it is becoming increasingly important to evaluate their performance. The current paper introduces a novel statistical me...
Abstract. We present an implementation of model-based online reinforcement learning (RL) for continuous domains with deterministic transitions that is specifically designed to achi...
— In many visual multi-object tracking applications, the question when to add or remove a target is not trivial due to, for example, erroneous outputs of object detectors or obse...