Facial Animation Design of Chinese Shaanxi Xi'an Dialect Based on Dominance Model in Speech Assistant System

This paper presents a comprehensive virtual animated articulation model based on the dominance concept of the Shaanxi Xi'an dialect, designed for integration into Speech Assistant Systems (SA) to support deaf-mute students and individuals learning this dialect. A detailed multimodal database has been developed, capturing the three-dimensional movements of the tongue, lips, and jaw for both vowels and consonants, with precise temporal resolution. The model not only represents isolated phonemes but also systematically incorporates the interactions between consecutive phonemes, articulatory configurations, and dynamic timing patterns, enabling realistic simulation of natural speech production. Advanced algorithms are employed to synchronize articulatory movements with speech sounds, ensuring real-time animation that accurately reflects subtle variations in pronunciation and coarticulation effects. This approach allows learners to visualize the complete articulatory process, providing an intuitive and interactive tool for language acquisition, speech training, and therapy applications. Beyond the immediate educational context, the methodology offers a scalable framework for extending real-time articulation animation to other languages, supporting cross-linguistic studies of phonetic articulation, enhancing human-computer interaction, and improving accessibility in speech-based systems. The proposed model thus combines linguistic theory, computational modeling, and animation techniques to deliver both practical educational tools and a foundation for further research in speech visualization technologies.