automatic lip reading

automatic lip reading

Understanding speech from the images of a person's lip movements. Automatic lip reading (ALR) converts lip patterns (visemes) into sounds (phonemes) that are turned into words. The first ALR is attributed to Frank Hubner, who in 2006, was able to reconstruct much of the dialog in several of Hitler's home movies, taken at his mountain retreat in the Bavarian Alps during World War II.

Although automatic lip reading is envisioned to provide an extra degree of surveillance capability in the future, it is only successful if the speaker is directly facing the camera. In addition, the slower the speech, the more effective the results.
References in periodicals archive ?
Petajan, who is highly regarded in the technical community for his work in digital video compression, has dedicated the past seventeen years to solving the automatic lip reading (ALR) problem associated with animation.
Full browser ?