Mouth detector with combination between LipNet and PyImageSearch

Le Minh Quan Tran 1, * and Linh Nguyen Hoang Anh 2

1 Mechatronic Systems Engineering, Faculty of Technology and Bionics, Hochschule Rhein-Waal, Germany.
2 Information Technology, Faculty of Interdisciplinary Science, Vietnam National University Ho Chi Minh City - University of Science, Vietnam.
 
Research Article
World Journal of Advanced Research and Reviews, 2024, 22(03), 065–073
Article DOI: 10.30574/wjarr.2024.22.3.1633
Publication history: 
Received on 17 April 2024; revised on 29 May 2024; accepted on 01 June 2024
 
Abstract: 
In the digital age, the authenticity and integrity of multimedia content have become increasingly critical, especially in contexts involving high-profile political figures like former President Donald Trump. The proliferation of manipulated and deepfake videos poses significant threats to public discourse and trust. This project addresses these challenges by developing a lip-reading system using PyImageSearch and Python TensorFlow to verify and validate spoken content in videos. By leveraging advanced computer vision techniques and machine learning models, the system aims to analyze and interpret lip movements accurately, ensuring the correspondence between the spoken words and the audio track. This tool can be instrumental in detecting inconsistencies and potential manipulations in videos, thereby enhancing the reliability of digital content. The ultimate goal is to provide a robust solution that can be utilized by media organizations, fact-checkers, and the general public to maintain the integrity of political communications in the digital realm.
 
Keywords: 
Mouth Open Detector; Facial Recognition; PyImage Search; LipNet
 
Full text article in PDF: 
Share this