Published Paper Details:

The Human-Computer Interaction application discussed in this research paper focuses on cursor control using facial movements captured through a standard webcam. This innovative system operates without the need for any wearable hardware or sensors, relying solely on the analysis of the user's facial gestures. The core functionality revolves around predicting facial landmarks, particularly eye blinks and mouth movements, within a video stream. Various actions correspond to specific facial movements: opening the mouth activates or deactivates mouse control, winking the right eye triggers a right-click function, winking the left eye initiates a left-click action, squinting detects the activation or deactivation of scrolling, and head movements (pitch and yaw) control scrolling or cursor movement. The system relies on facial landmarks to derive features that enable the detection of these actions. For instance, eye aspect ratio (EAR) helps identify winks and blinks, while mouth aspect ratio (MAR) indicates whether the mouth is open or closed, essential triggers for mouse control. This model primarily focuses on analyzing two main factors: Eye Aspect Ratio (EAR) for detecting winks and blinks, and Mouth Aspect Ratio (MAR) to determine the status of the mouth--whether it's open or closed. These ratios serve as crucial indicators for interpreting various facial gestures that, in turn, trigger specific mouse control actions.