This repository contains the code for Human Face Landmark Detection using Landmark Guided Face Parsing (LaPa) dataset. The pre-trained MobileNetv2 is used for the task in the TensorFlow framework.
YouTube tutorial: Human Face Landmark Detection in TensorFlow using MobileNetv2
The LaPa dataset contains the training, validation and testing dataset. Each dataset have images, segmentation mask and the 106 human facial key points.
Download the dataset: Landmark Guided Face Parsing (LaPa)
Dataset paper: A New Dataset and Boundary-Attention Semantic Segmentation for Face Parsing
The first image has Ground-Truth Landmarks and the second image has Predicted Landmarks.
For more follow me on: