Speech driven 3d facial animation lacks natural lip movement shape
The Shape of Speech: A Geometric Measure of Coarticulation for Speech-Driven 3D Facial Animation
GraphicsComputer Vision and Pattern RecognitionMultimedia
Summary
When computers create 3D talking faces, they often miss the subtle way lips move between sounds, called coarticulation. This paper introduces a new way to measure how lips move during speech by comparing the path lips take with the shortest possible path through key mouth shapes. The authors tested four top animation systems and found they all make lip movements simpler and less natural than real speech. They also ran a study showing people prefer faces that keep these natural lip movements. This new measure helps pinpoint how to improve computer-generated talking faces to look more realistic.
What this means in practice
- •For game developers: Improve character lip-sync animations to capture more natural speech movements using the new geometric lip-path measure.
- •For animation studios: Evaluate and enhance 3D facial animations for films by identifying and restoring lost natural lip trajectory details.
Authors
Danzel Serrano, Przemyslaw Musialski
Abstract
Speech-driven 3D facial animation can reproduce recognizable mouth poses. However, it can simplify the motion between them, and that motion carries coarticulation, the way the sounds around each sound shape its articulation. We introduce a geometric measure of this trajectory shaping: lip-path length compared with the shortest route through the vowel, consonant and vowel positions of a speech segment. In contrast to the endpoint chord, this consonant-aware route accounts for obligatory transit and avoids degeneracy, while preserving invariance to uniform motion gain. The measure needs only a forced alignment, so it applies where no ground truth exists. We demonstrate it on four state-of-the-art methods, one per architectural family, real-time and offline. All four trace flatter lip trajectories than captured speech. Against frame-rate-matched ground truth, DiffPoseTalk, ARTalk and FaceFormer show clear deficits, equivalent on this measure to removing 15-60% of real speech's fast articulatory component. CodeTalker is marginal on the primary measure and clear on a companion measure. A pre-registered study with 97 viewers and 3,523 judgments underpins the measured direction: controlled damping of real motion lowers the score and is penalized, whereas exaggeration shows no detected penalty over the tested range. Viewers also prefer real speech in 73.4% of sentence comparisons and, in the aggregate, on single words. Together, the measure, its calibration and the study identify a perceptually relevant loss of trajectory shaping and a concrete target for improving synthesized articulation.