Aimmia Virtual Try-On runtime assets

MediaPipe Tasks Vision 1.0.1, Copyright Google LLC / The MediaPipe Authors.
JavaScript and WebAssembly runtime: Apache License 2.0 (see LICENSE.txt).
https://github.com/google-ai-edge/mediapipe
https://www.npmjs.com/package/@mediapipe/tasks-vision/v/1.0.1

Face Landmarker model bundle:
https://storage.googleapis.com/mediapipe-models/face_landmarker/face_landmarker/float16/1/face_landmarker.task
Model documentation:
https://ai.google.dev/edge/mediapipe/solutions/vision/face_landmarker

These assets run locally to locate facial landmarks for cosmetic colour rendering.
They do not identify the visitor or diagnose skin conditions.

Pose Landmarker Lite model (float16, version 1), Google MediaPipe:
https://storage.googleapis.com/mediapipe-models/pose_landmarker/pose_landmarker_lite/float16/1/pose_landmarker_lite.task
https://ai.google.dev/edge/mediapipe/solutions/pose_landmarker
Used locally to track shoulders for the two neck-jewellery collections.
The existing BiSeNet model also segments the visible neck; see
../ear-parser/NOTICE.txt for its source and licence.

sample-neckline.webp is an Unsplash portrait with the neck/shoulders in view.
Source: https://images.unsplash.com/photo-1534528741775-53994a69daeb
Licence: https://unsplash.com/license
Used for the neckline sample only; no endorsement is implied.

The default sample-portrait.webp is a cropped and resized stock photograph from Unsplash.
Source: https://images.unsplash.com/photo-1524504388940-b1c1722653e1
Licence: https://unsplash.com/license
The photograph is used as an illustrative sample; no endorsement is implied.

The earlier demo-portrait.webp is a render of the CC0 MPFB/MakeHuman model and is
no longer the default. See ../face/FEMALE_MODEL_LICENSE.txt for that asset's source.