This is a Tensorflow 2 implementation of the paper HyperFace: A Deep Multi-task Learning Framework for Face Detection, Landmark Localization, Pose Estimation, and Gender Recognition (IEEE TPAMI 2016).
The main idea of the paper is to fuse the intermediate layers of a deep CNN in order to train different tasks under consideration. It presents an architecture to learn common features for these tasks and exploit the synergy among them.
The following are model's predictions for image descriptions:
