Voice & Gesture-Controlled File Management Using Computer Vision

In today's computing environments, user interaction techniques that are both accessible and secure are essential. Using Passwords and other traditional security measures for managing the files are susceptible to attack. Problem and issues may arise by continuing to rely on keyboard and mouse inputs. In order to provide secure document access, this paper proposes a multi-modal biometric framework that combines gesture-based control and voice authentication. A Support Vector Machine (SVM) classifier used in the system's authentication component was trained using Mel-Frequency Cepstral Coefficients (MFCCs) extracted from user voice inputs. This aids in effectively distinguishing between authorized and unauthorized users. To increase the resilience of the model, data augmentation techniques such as pitch modulation and noise injection are used. After voice authentication is successful, a gesture recognition module is launched. This uses OpenCV and Google's MediaPipe framework to track hand motion in real time. This feature enables the user to move through files with simple hand gestures. An authenticated user is granted access to a secure folder by the assistant. The speech recognition library processes voice commands that allow the user to choose files and move through them. This study presents a robust and user-friendly file assistant that uses voice biometrics to increase security while also improving operational efficiency with an easy-to-use gesture-based interface.