Article: Three-dimensional facial image modelling and animation generation method integrated with emotion recognition Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.52 pp.117 - 134 Abstract: This study proposes a method for 3D facial modelling and animation generation integrated with emotion recognition. This method deeply combines high-precision emotion recognition with animation driving to improve the naturalness of facial expressions, emotional accuracy, and real-time interaction capabilities. Experimental results show that the average accuracy rate of 3D facial emotion recognition increases from 89.1% to 92.1%, with happiness (HA) reaching 98.9%, verifying the model's high reliability and stability. In animation generation experiments, the emotion recognition accuracy reaches 90%, emotional consistency is 88%, and the average frame generation time is 48 ms/frame, all outperforming the control model. The research innovation and contribution lie in proposing a systematic integration strategy of emotion recognition with 3D modelling and animation generation. This enriches the theoretical framework of facial animation generation, achieving a balance among accuracy, naturalness, and efficiency, and providing an efficient and feasible technical solution. Inderscience Publishers - linking academia, business and industry through research

Title: Three-dimensional facial image modelling and animation generation method integrated with emotion recognition

Authors: Xiaowen Guo

Addresses: College of Art and Creativity, Anhui University of Applied Technology, Hefei, 230000, China

Abstract: This study proposes a method for 3D facial modelling and animation generation integrated with emotion recognition. This method deeply combines high-precision emotion recognition with animation driving to improve the naturalness of facial expressions, emotional accuracy, and real-time interaction capabilities. Experimental results show that the average accuracy rate of 3D facial emotion recognition increases from 89.1% to 92.1%, with happiness (HA) reaching 98.9%, verifying the model's high reliability and stability. In animation generation experiments, the emotion recognition accuracy reaches 90%, emotional consistency is 88%, and the average frame generation time is 48 ms/frame, all outperforming the control model. The research innovation and contribution lie in proposing a systematic integration strategy of emotion recognition with 3D modelling and animation generation. This enriches the theoretical framework of facial animation generation, achieving a balance among accuracy, naturalness, and efficiency, and providing an efficient and feasible technical solution.

Keywords: emotion recognition; 3D facial image; facial modelling; animation generation.

DOI: 10.1504/IJICT.2025.151168

International Journal of Information and Communication Technology, 2025 Vol.26 No.52, pp.117 - 134

Received: 29 Sep 2025
Accepted: 31 Oct 2025
Published online: 15 Jan 2026 *

Title: Three-dimensional facial image modelling and animation generation method integrated with emotion recognition

Keep up-to-date