Bilkent University
Department of Computer Engineering


OCEAN aware Conversational Virtual Avatar Behavior based on Laban Movement Analysis


Sinan Sonlu
MS Student
Computer Engineering Department
Bilkent University

High cost of case specific animation creates a demand on procedural methods for generating and altering 3D articulated human animation in video games, virtual tutoring systems and interactive social environments. Altering a base animation according to the content of the conversation requires a psychological processing of the used natural language. OCEAN Personality Parameters are used for defining the character of an individual. Using an input dialogue text and facial expression, OCEAN Parameters are found and mapped to Laban Effort Parameters. Laban Movement System introduces a methodology to alter animation, so that the result would appear to express one or more orthogonal movement features. Output animation is realized using Inverse Kinematics, with the help of multiple IK targets around the 3D model, based on Laban Shape Qualities. The resulting system makes it possible to alter the behavior of an OCEAN aware Virtual Avatar, using dialogue and facial expression.


DATE: 19 March, 2018, Monday, CS590 & CS690 presentations begin at @ 15:40