Face Gesture Stylizer: Real-Time Facial and Hair Restyling Using Hand Gesture Recognition and Computer Vision
Year: 2026
Type: article
Abstract: Face Gesture Stylizer is a webcam-based computer-vision application that applies artistic transformations to a user’s face and hair while retaining the natural appearance of the hands, neck, clothing and background. The application uses hand gestures to control the live effect: a right or left swipe changes the selected style, an upward or downward swipe adjusts blend intensity, and a sustained thumb–index pinch requests a photograph capture. The browser workflow uses MediaPipe Tasks Vision for hand landmark detection and selfie multiclass segmentation. A gesture engine interprets landmark movement over time, while a compositor applies the selected effect through an estimated head mask and preserves the hand region as a foreground object. The interface provides 24 built-in artistic styles and an optional FLUX.2 [klein] generative-editing path when a compatible backend is configured. The supplied screenshots demonstrate the live interface and three selected styles. Instantaneous frame-rate readings of 15–19 FPS are visible in the captures; these are observations rather than controlled benchmark results.
Authors: Puja Gupta, Srijan Mohan, Rishab Patidar
Language: English
Subfield: Human-Computer Interaction
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225392715