README

Introduction

Here we provide some lightweight facial landmark models with fast coordinate regression. The input of these models is loose cropped face image while the output is the direct landmark coordinates.

Pretrained Models

• Model 2d106det

  2021.07: We now support model inference by our insightface python package, please check image_infer.py for detail.

  Given face detection bounding box, predict 2d-106 landmarks. Mainly used for static image inference.

  Backbone: MobileNet-0.5, size 5MB.

  Input: size 192x192, loose cropped detection bounding-box.

  Download link:

  baidu cloud (code: lqvv)

  google drive

• Model 2d106track

  Given landmarks bounding box, predict 2d-106 landmarks. Used for video landmarks tracking.

  Download link: coming soon

Visualization

<p align="center">Points mark-up(ordered by point names):</p> <div align="center"> </div> <p align="center">Image result:</p> <div align="center"> </div> <p align="center">Video result:</p> <div align="center"> </div>

FAQ