People are striving to teach a variety of computing devices things like recognition of images, the ability to navigate in space, understanding human speech. Some of these problems are being successfully implemented by technical means.
Computer can recognize the voice. In this is sure Ahmad Hassanat , a researcher from Jordan. He explains that human speech is used in more than 50 different sounds, while for their casting lips may take only 10-14 forms. Thus, lip reading requires not only a comparison of the shape and sound, but also to monitor the "body language", knowledge of the context of conversation and intuition.
Experiments show that the machine lip-reading, even from a limited set of words (such as numbers) is successful only in 50% of cases, which is not a satisfactory result.
Visual speech recognition identifies a number of technical problems and raises questions about the confidentiality and integrity of personal space.