Lip Animation Synthesis Using Phonetic Symbol Matching

The patent introduces an intermediary lip animation synthesis system that mediates between the speech output and the visual display. This system uses phonetic symbol databases and motion synthesis algorithms to generate realistic lip movements that correspond to the spoken words, rather than using simple mouth opening and closing animations. The intermediary system processes phonetic information and transforms it into corresponding visual lip motions, thereby improving animation realism without requiring complex direct control mechanisms.

The patent applies preliminary action by pre-building comprehensive phonetic-symbol lip-motion matching databases before actual speech animation is needed. During system operation, the database simply needs to retrieve pre-computed lip motion sequences corresponding to phonetic symbols, rather than computing animations in real-time. This pre-processing approach stores complex motion data in advance, allowing the system to quickly generate diverse and realistic lip animations during actual speech without requiring complex real-time computation.

The patent uses preliminary action by pre-computing and storing lip motion sequences in the phonetic-symbol lip-motion matching database before actual speech synthesis is needed. During real-time operation, the system retrieves pre-computed motion sequences based on phonetic symbols and timestamps, rather than computing synchronized animations in real-time. This approach maintains high synchronization accuracy between speech and lip movements while significantly reducing processing time during actual speech output.