Face landmark localization with single deep neural network: https://arxiv.org/abs/1702.02719v1
Facial Landmark Detection by Deep Multi-task Learning: http://mmlab.ie.cuhk.edu.hk/projects/TCDCN.html
A novel face alignment method using single deep network (SDN) on existing limited training data. Rather than using a max-pooling layer followed one convolutional layer in typical convolutional neural networks (CNN), SDN adopts a stack of 3 layer groups instead. Each group layer contains two convolutional layers and a max-pooling layer, which can extract the features hierarchically. Moreover, an effective data augmentation strategy and corresponding training skills are also proposed to over-come the lack of training images on COFW and 300-W da-tasets. The experiment results show that our method outper-forms state-of-the-art methods in both detection accuracy and speed.
PDF: https://arxiv.org/pdf/1702.02719v1
Facial landmark localization – Dataset and Code
1. Face landmark localization demo: [download]
2. Multi-Task Face Landmark localization dataset: [download] (12,995 face images)
3. Multi-Attribute Facial Landmark Localization (MAFL) dataset: [download] (20,000 face images)
4. TCDCN face alignment tool – Face landmark localization binary:
It takes an face image as input and output the locations of 68 facial landmarks. Win32 Binary [download] Matlab [download]
