Engineering an Illusion

At Speech Graphics we specialize in one of the holy grails of computer facial animation: automatic, accurate lip sync. Given audio, our task is to create the illusion that the animated face you see is the source of the sound you hear. This illusion is notoriously difficult to achieve: the movements of speech are fast, complex and subtle, and viewers are highly sensitive to any mismatch between the face and the voice. Our world-class pipeline incorporates powerful speech analysis and procedural animation techniques to achieve this illusion consistently, allowing animators to handle overwhelming demand for high-quality dialogue.

But Speech Graphics goes beyond good lip sync. Speech contains energy and emotion, and that too can be decoded from the voice and synchronized in the face. Using all available acoustic information, our algorithms drive not just the mouth but the entire face from audio input, from syllables to scowls.

This is an interdisciplinary problem and Speech Graphics has a unique fusion of expertise in speech technology and computer animation found nowhere else in the world. Our team brings together decades of research into phonetics, task dynamics, speech recognition, psychology, machine learning, computer graphics, physical models and more. And we have solved the problem of how best to apply our systems to real-world facial rigs.