Volume : VIII, Issue : VIII, August - 2019

Lip reading through fourier based facial landmark

Nandini M S, Nagappa U. Bhajantri, Trisiladevi C Nagavi

Abstract :

An algorithm has been presented to reduce noise distortion and to predict the movement of lips under noisy environment. However, such environment will be a challenging problem for a good heå person. Further it is very difficult to listen and understand the words spoken by a person under noise distorted conditions. But, heå impaired person reads the movement of lips by analyzing the shapes of lips provided language is known . The work is proposed to detect, track and recognize the shapes of lips and predict the movement of lips for Kannada language. The lip tracking, and prediction of sentences spoken by a person, as we have to keep track of every changes that are observed in lip movements recognized at a fraction of seconds measured in terms of time. Thus, time is a parameter that plays a vital role in analyzing and understanding the sentences spoken by a person. Further effort attempt to perform the task of estimating the shape of lips and then annotates the shapes of every lip movement with appropriate words. A minute change in movement of lips designates a specific prediction of words spoken. As the movement of tongue cannot be seen in videos properly, every change in shape of a lip designates a specific words of a sentences. Even while recognizing specific words like Amma and Appa are almost similar in nature even though they are different in meaning. Thus, we have came up with a model that analyzes and predicts the language spoken by a person with a good accuracy

Keywords :

Article: Download PDF   DOI : 10.36106/ijsr  

Cite This Article:

LIP READING THROUGH FOURIER BASED FACIAL LANDMARK, Nandini M S, Nagappa U.Bhajantri, Trisiladevi C Nagavi INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH : Volume-8 | Issue-8 | August-2019

Number of Downloads : 553

References :