Many statistics have confirmed that many deaf are enabled to access to written information. As a solution, computer applications designed for deaf person, have been created. Therefore, to facilitate access to information, new methods improving the dialogue between human and machine are required. The signs generation is based on different parameters such as manual configuration, orientation of hands, the location where the sign is made, the movement made by hand and the facial expression accompanying the realization of the sign. We take into account all these parameters and the system presented in this paper is based also on avatars which have many degrees of freedom. The challenge of this project is to find the tradeoff between computational time and realistic representation that must be closer to real-time generation signs.
展开▼