lipreading

Lipreading, also known as speechreading, is the process of understanding spoken language by interpreting visual information produced by the speaker's mouth, facial expressions, and gestures. It does not require auditory input and is often used by deaf and hard-of-hearing individuals, but many people use lipreading to supplement auditory speech in noisy environments.

Lipreading relies on a combination of cues: the shapes and movements of the lips, teeth, and tongue; the speaker's jaw position; eye gaze; facial expression; and contextual information such as syntax and topic.

Some phonemes are visually indistinct; for example, many vowels and consonants share similar lip movements, and complete intelligibility depends on timing, coarticulation, and context. Lighting, view angle, occlusions (glasses, facial hair, masks), speech rate, and accents can further affect accuracy. Because of these limits, lipreading is usually not sufficient to understand unfamiliar conversations without context, and many conversations remain partly unintelligible.

People can improve lipreading skills through training that teaches strategies for anticipating words, using context, and recognizing distinctive visual patterns. Programs are used in education for deaf and hard-of-hearing students, in professional interpreting, and in everyday communication.

Technological research includes automatic lipreading systems based on computer vision and machine learning. While such systems show progress, they are not yet reliable in general use and are typically used as assistive tools or for research, with privacy and consent considerations.