Face Quality - 3.5 English

Vitis AI Library User Guide (UG1354)

Document ID
Release Date
3.5 English

The Face Quality library uses the face quality network to detect the quality score of a face. If a person is facing the camera with no obstructions, the score is high. On the contrary, a blurry or face in profile will get a low score. The scores range from 0 to 1. It also provides face landmark positions. The input is a face that is detected by the face detect network and the output contains a quality score and five landmark key points.

Figure 1. Face Quality Example

The following table lists the face quality models supported by the Vitis AI Library.

Table 1. Face Quality Models List
No Model Name Framework
1 face-quality Caffe
2 face-quality_pt PyTorch