ANALYZING THE QUALITY OF THREE-DIMENSIONS CHARACTER ANIMATIONS COMPOSED BY ANN

Abstract

Background problem: ANN’s use to produce character animation has grown speedily in recent years. Nonetheless, the perceived realism and quality comparison of this new approach are less if not negligible compared to traditional pipe results. Main Objective: The purpose of this research is to present the preliminary data in case the visual character of this novel approach is considered to be a greater visual character than an identical keyframe-based approach, and to generate animations as a basis for identification for that approach. Research Method: “Keyframe animations” that perform actions identical to “Artificial Neural Network-based plug-ins” for unified game engines were created by hand following a questionnaire study. This research was appointed to different accomplished and undisciplined willing participants to indicate civil assumption. Finding/Result: The outcome of this study indicates that participants perceive ANN-based plug-in animation as a whole to be extra realistic, essential, steady, and attractive. Analytical investigation via “t-test” showed high analytical implications when measuring assumptions about character between the 2 groups. Conclusion: The “Artificial Neural Network-based plug-ins” approach was treated by the entries to have an exceptional visual character for the argument mentioned raised. All competent entries accurately estimated by the 2 animation sets were “Artificial Intelligence” based and a third of the incompetent ones. Nevertheless, the incompetent entries that estimated incorrectly expressed the same motivations for their guesses as those who guessed correctly.