The generation and recognition of body language is a key technologies of VR. Sign Language is a visual-gestural language mainly used by hearing-impaired people. In this paper, gesture and facial expression models are created using computer graphics and used to synthesize Chinese Sign Language (CSL), and from it a human-computer CSL interaction system is implemented. Using a system combining CSL synthesis and CSL recognition subsystem, hearing-impaired people with data-gloves can pantomime CSL, which can then be displayed on the computer screen in real time and translated into Chinese text. Hearing people can also use the system by entering Chinese text, which is translated into CSL and displayed on the computer screen. In this way hearing-impaired people and hearing people can communicate with each other conveniently.