A Multi-Stage Approach to Facial Feature Detection

• DOI: 10.5244/c.18.30
• Publication Date: 2012-02-07T13:11:18Z
• Authors: D. Cristinacce; T. Cootes; I. Scott
• Keywords: Feature (linguistics)

We describe a novel shape constraint technique which is incorporated into multi-stage algorithm to automatically locate features o n the human face. The method coarse-to-fine. First face detector applie d find approximate scale and location of in image. Then individual feature detectors are applied combined using known as Pairwise Reinforcement Feature Responses (PRFR). points predicted by this then refined version Active Appe arance Model (AAM) search, tuned edge corner features. final output three stage shown give much better results than any other combination methods. outperforms previous published on BIOID test set [11].