Autonomous Emotional Virtual Character: An Approach with Deep and Goal-Parameterized Reinforcement Learning

Abstract

We have developed an autonomous virtual character guided by emotions. The agent is a virtual character who lives in a three-dimensional maze world. We found that emotion drivers can induce the behavior of a trained agent. Our approach is a case of goal parameterized reinforcement learning. Thus, we create conditioning between emotion drivers and a set of goals that determine the behavioral profile of a virtual character. We train agents who can randomly assume these goals while trying to maximize a reward function based on intrinsic and extrinsic motivations. A mapping between motivation and emotion was carried out. So, the agent learned a behavior profile as a training goal. The developed approach was integrated with the Advantage Actor-Critic (A3C) algorithm. Experiments showed that this approach produces behaviors consistent with the objectives given to agents, and has potential for the development of believable virtual characters.

3D Modeling Design of Multirole Virtual Character Based on Visual Communication in Wireless Sensor Networks

“…The comparison methods used include the character virtual modeling design method based on VR technology proposed in Literature [8], the character virtual modeling design method based on parameter mode SMPL proposed in Literature [9], the character virtual modeling design method based on the mobile platform proposed in Literature [10], the character virtual modeling design method based on a computer diagram proposed in Literature [11], and the character virtual modeling design method based on deep and enhanced target parameter learning proposed in Literature [12]. According to Figure 5, there are different degrees of gaps in the data coverage of the five traditional design methods, which shows that there are data leakage points in these methods when scanning and positioning human structure (a) Literature [8] method (b) Literature [9] method (c) Literature [10] method (d) Literature [11] method (e) Literature [12] method (f) The proposed method Literature [8] algorithm Literature [9] algorithm Literature [10] algorithm Literature [11] algorithm Literature [12] algorithm The proposed algorithm 7 Wireless Communications and Mobile Computing information, which affects the subsequent generation effect of virtual character 3D modeling. The data coverage of this method is not blank, which shows that the scanning data obtained under the support of a wireless sensor network positioning algorithm is complete and can cover the whole scanning area, which provides more data support for generating virtual character 3D modeling.…”

“…Literature [11] proposed a character virtual modeling design method based on the computer diagram, in which the real-time computer diagrammatic technology, 3D modeling technology, and double-eyed 3D virtual technology are used for the modeling design of various virtual characters in virtual reality technology. Literature [12] proposed a character virtual modeling design method based on deep and enhanced target parameter learning, in which the maximum rewarding function of character virtual modeling design is constructed, and the mapping relationship between different kinds and various characters is studied. Besides, the agent learning method is used to complete the character virtual modeling design.…”

“…(a) Literature[8] method (b) Literature[9] method (c) Literature[10] method (d) Literature[11] method (e) Literature[12] method (f) The proposed method…”