yemount/pose-animator
Pose Animator
Pose-animator is a system that maps real-time body and face tracking data to 2D vector illustrations. It functions as a skeletal animation engine and motion controller that translates human keypoint recognition into instantaneous SVG path updates.
The project enables real-time motion capture from webcam feeds and pose extraction from static images. It utilizes a skeletal rig to link virtual bones to vector character surfaces, allowing for the animation of custom characters and interactive avatars.
The tool incorporates client-side machine learning inference for processing camera frames, coordinate space mapping, and frame interpolation to smooth character motion. It supports the import of custom character rigs via SVG files containing defined skeleton and illustration groups.
Features
- Real-Time 2D Avatar Animation - Provides real-time mapping of human body and face movements to a 2D vector character avatar.
- Client-Side Inference - Runs PoseNet and FaceMesh models directly in the browser for local processing of camera frames.
- Vector Animation Libraries - Animates 2D vector illustrations by linking their paths to a skeletal system for fluid movement.
- Skeletal-Driven Vector Animations - Implements skeletal animation to map human movements onto 2D vector illustrations.
- Keypoint-Driven Path Transformations - Updates SVG curve coordinates by calculating point offsets and rotations based on detected body landmarks.
- 2D Skeletal Rigs - Functions as a skeletal animation engine that moves vector character surfaces via a virtual rig.
- 2D Skeletal Vector Animation - Maps PoseNet keypoints to a virtual bone hierarchy that transforms linked vector paths in real time.
- Pose-Driven Character Rigging - Uses PoseNet and FaceMesh to drive the movements of a custom vector character based on human keypoints.
- Pose-to-Vector Animation Tools - Maps real-time body and face tracking data from PoseNet and FaceMesh to 2D vector illustrations.
- Real-Time Motion Tracking - Processes live webcam streams to recognize keypoints that drive character animations.
- Real-Time Motion-Driven Avatars - Creates digital avatars that mirror a user's real-time physical expressions and gestures.
- Keypoint-Reactive SVG Path Updates - Updates SVG paths instantaneously based on human keypoint recognition in the browser.
- Single-Image Pose and Expression Inference - Analyzes a single static image to extract a person's position for application to a vector character.
- Motion-Based Frame Interpolation - Smooths the transitions between disparate ML detection results to prevent jittering in character motion.
- SVG Group Rigging - Uses nested SVG group structures to define the parent-child relationships between bones and the character surface.
- Vector Character Rig Importers - Supports the import of custom character rigs via SVG files containing defined skeleton and illustration groups.
- Webcam Pose Tracking Interfaces - Translates live camera feeds into character animations through real-time motion capture.
- Coordinate Space Mappers - Translates raw camera pixel coordinates into a normalized coordinate system for the vector canvas.
