A brief history of lip-reading:
Although lip-reading (also called speech-reading) is done subconsciously since the very beginning of human existence. It's been particularly put into use for more helpful reasons. Mainly in communication strategies, as in aiding hearing-impaired and deaf people in communicating with the world and also helping normal people recognize words in noisy environments. However, it is very likely that the best lip reader can only catch 50% of the words uttered. This caused Lip-reading to become a hot topic in many scientific fields "e.g. medicine, psychology, psycholinguistics and especially technology". Recent researches have shown that in automatic lip-reading systems there is a chance to achieve up to 55 % error reduction, and the low recognition rate in the acoustic only system (43%) is raised to 75% by the incorporation of visual information.
Modern Computer-based lip-reading usages:
In addition to speech-recognition, speech impaired aid and increasing robustness in noisy places; modern Computer-based lip-reading technology are also used for security issues, and would even help video surveillance systems spot people planning a crime or terror attack!
In this project, we will use Computer-based lip-reading technology in increasing the level of information security (specifically of passwords, PIN, etc).
According to Roland Goecke (1998), current speech recognition systems use powerful statistical models of the audio components of spoken language but can fail unpredictably in non-ideal acoustic conditions. This in addition to the short period of time caused us to limit our experiment on basic digits only (from 0-9), presuming ideal conditions. The goal of our work is to develop a speech-reading system that helps in preventing the current major security issue of shoulder-surfing, which is a common, affective type of attacks that involves using direct observation techniques, by either distracting the...