Quarterly Scientific Journal of Audio-Visual Media

Quarterly Scientific Journal of Audio-Visual Media

Creating Intelligent and Self-Aware Characters in Animations Using Deep Learning

Document Type : Research Paper

Author
Ahmad Soflaei
Assistant Professor, Department of Digital Arts, Faculty of Television and Digital Arts, IRIB University, Tehran, Iran
10.22085/javm.2025.543254.2366
Abstract
The present article aims to examine the role of deep learning in creating intelligent and self-aware characters in animations. The research method is descriptive-analytical, using previous data and analyzing deep learning patterns. Techniques such as artificial neural networks (ANNs), convolutional (CNNs), recurrent (RNNs), and GAN patterns were applied to design characters capable of regulating behavior, making independent decisions, and continuous learning. The findings show that these technologies have increased the realism and interaction of animated characters through reinforcement learning, behavior analysis, and multimodal data processing. However, ethical and philosophical challenges such as the issue of virtual character rights, the impact on traditional artists, and technological abuses require serious attention and consideration.
Keywords
Deep Learning, Intelligent Characters, Self-Aware Characters, Artificial Intelligence, Animation