Voice Animator: Automatic Lip-Synching in Limited Animation by Audio

Abstract

Limited animation is one of the traditional techniques for producing cartoon animations. Owing to its expressive style, it has been enjoyed around the world. However, producing high quality animations using this limited style is time-consuming and costly for animators. Furthermore, proper synchronization between the voice-actor’s voice and the character’s mouth and lip motion requires well-experienced animators. This is essential because viewers are very sensitive to audio-lip discrepancies. In this paper, we propose a method that automatically creates high-quality limited-style lip-synched animations using audio tracks. Our system can be applied for creating not only the original animations but also dubbed ones independently of languages. Because our approach follows the standard workflow employed in cartoon animation production, our system can successfully assist animators. In addition, users can implement our system as a plug-in of a standard tool for creating animations (Adobe After Effects) and can easily arrange character lip motion to suit their own style. We visually evaluate our results both absolutely and relatively by comparing them with those of previous works. From the user evaluations, we confirm that our algorithms is able to successfully generate more natural audio-mouth synchronizations in limited-style lip-synched animations than previous algorithms.