Face Rig & Animation
Frontal head render → soft facial masks, 2D landmarks, and region scores for lip sync and expression in the AutoRig pipeline.
Head analysis
Soft masks
LandmarksRuns the same analyze-head call as the worker tester.
Pipeline
Service ingests RGB plus synthetic depth (luminance) and alpha, like the pipeline tester. Output: region confidence, landmarks, contours, and soft masks for eyes, mouth, brows, etc.
Why masks
Masks delimit where mesh deformation and speech motion apply. Clean separation keeps talking, blinking, and emotion layers from fighting each other at runtime.
In AutoRig
After body rigging, the worker can analyze the head so you skip hand-painted weights per zone.
Demo uses POST …/api/face-rig/analyze-head on the live worker.
Interactive demo
Sample image runs on load. Replace the image, then Run analysis to refresh masks.
Drop or click to choose image
POST https://worker-0001.free3d.online/api/face-rig/analyze-head
Next steps
Add engine-specific notes (Unity blendshapes, Unreal poses), capture tips for stable front lighting, or troubleshooting when confidence is low. Well-lit, centered faces work best.