Viseme Morphing and Text-to-Speech Integration for Indonesian News Broadcasting

Advancements in multimedia technology and artificial intelligence have driven innovation in digital broadcasting, including virtual newsreaders. This study proposes a text-to-speech-based lip-sync animation system specifically for the Indonesian language to improve synchronization between lip movements and speech. The primary challenge in developing this system lies in generating realistic lip animations that correspond with the phonetic structure of Indonesian. The system workflow involves text input, syllable parsing using the Finite State Automata (FSA) method, viseme conversion (viseme morphing), and web-based animation output. Test results show a viseme duration accuracy of 98.5%, voice-lip movement synchronization of 94.26%, and a Mean Opinion Score (MOS) of 77.12%, indicating that the system is reasonably feasible for implementation. Despite minor delays, the system demonstrates strong potential for further development through the integration of Natural Language Processing (NLP) and deep learning, which could improve viseme mapping accuracy and enhance system flexibility across various digital broadcasting platforms.