Frame the avatar head and shoulders on the video stage
main·
Commit

The package draws the avatar life-size, so the README camera (0.15 m high, 2.2 m back) only showed the legs and waist in our portrait stage. HeadShot measures the model once it is in the scene and points the camera at the head and shoulders, fitting both height and width, and reframes when the stage resizes (off stage on setup, then beside the interview, and on phones). Dev-only preview at ?dev=avatar (loadModels off, avatar file only).
Entire-Checkpoint: 01M41V4T1EHEHMKGK6XHXSQM0M
Checkpoints
Checkpoint 1