Problem: Developing algorithms of hand gesture recognition that are then integrated into systems supporting the communication in hard-of-hearing community.
First approach proposed a feature named Rank-Order Correlation Matrix that was extracted on static hand gestures represented by depth images.
Paper (ATC, IEEE)
Second approach used geometry-based features computed from silhouette to classify static hand gestures.
Paper (ICARCV, IEEE)
Third approach modeled dynamic hand gestures using geometrical characteristics extracted from hand silhouettes.
Paper (ATC, IEEE)