Face Landmark Detection - 1.1 English

Vitis AI Library User Guide (UG1354)

Document ID
Release Date
1.1 English

The Face Landmark Network is used to detect five key points of a face. The five points include the left eye, the right eye, nose, left lip of mouth, right lip of mouth. This network is used to correct face direction before face feature extraction. The input image should be a face which is detected by the face detection network. The outputs of the network are 5 key points. The 5 key points are normalized. The following image show the result of face detection.

Figure 1. Face Landmark Detection Example

The following table shows the face landmark model supported by the AI Library.

Table 1. The face landmark model list
No Model Name Framework
1 face_landmark Caffe