Clmtrackr

clmtrackr is a JavaScript computer vision library designed for facial landmark detection and real-time tracking. It implements Constrained Local Models to identify specific coordinate points on a human face within video feeds or static images.

The project functions as a real-time face warping engine and expression analysis tool. It can distort facial images via parametric models to create caricatures or identify and label emotional states such as happiness, sadness, anger, and surprise based on feature coordinates.

The library covers a broad range of capabilities including automatic and manual face detection, digital face masking, and image substitution. It provides tools for facial model rendering and visualization on a canvas, allowing for the overlay of graphics or the deformation of facial geometry in real time.

Features

Computer Vision Libraries - Provides a browser-based library for processing and analyzing visual data to track facial landmarks.

Facial Feature Tracking - Fits a deformable face model to a live video feed to identify and output specific facial landmark coordinates in real time.

Constrained Local Models - Implements Constrained Local Models to identify and track facial landmark coordinates in images and video.

Face Detection - Automatically locates human faces within image frames or video feeds to initialize tracking.

Facial Expression Recognition - Analyzes facial muscle movements and coordinates to automatically detect and label emotional states.

Facial Landmark Analysis - Identifies and follows the movement of specific facial landmarks in live video using a Constrained Local Model.

Facial Landmark Detection - Provides a JavaScript library for tracking facial landmarks and features in real time using Constrained Local Models.

Parametric Face Deformation - Alters facial appearance in real time using a parametric model with adjustable components.

Facial Graphic Overlays - Overlays digital graphics onto a video stream by fitting a facial model to features.

Emotion Analysis - Analyzes facial feature coordinates to identify and label emotional states such as happiness, sadness, anger, and surprise.

Image Warping Engines - Deforms pixels of a source image based on new facial coordinate points.

Real-Time Face Swapping Engines - Maps source facial features onto target subjects to substitute faces in real time.

Face Warping Engines - Implements a real-time engine for distorting facial images via parametric models.

Caricature Generation - Distorts facial coordinates based on a fit model to create a caricature effect in real time.

Coordinate-Based Warping - Transforms source images by mapping pixels between original coordinates and tracked facial landmarks to distort the face.

Facial Manipulation - Warps or distorts facial images and video in real time to create caricatures or alter appearance.

Facial Mesh Visualization - Renders tracked facial model data onto a canvas for real-time visual feedback.

Digital Face Filters - Overlays digital graphics or substitute images onto video streams by mapping them to tracked facial coordinates.

Blend Shapes - Alters facial geometry by manipulating weights associated with primary facial components to create expressions or deformations.

Facial Landmark Mapping - Identifies and maps the coordinate positions of facial landmarks within static images.

audunoclmtrackr

Features

Star history