Multiple Angles Face Detection

The video content presented here requires JavaScript to be enabled and the latest version of the Adobe Flash Player. If you are using a browser with JavaScript disabled please enable it now. Otherwise, please update your version of the free Adobe Flash Player by downloading here.


In this project, we merged three face detectors, one for frontal face, one for right profile face and one for left profile face, to a single detector. The three detectors were trained separately by using different training data, the algorithm of training is Adaboost. For multiple view face detection, the traditional way is to let the captured image go through the detectors trained for a certain degree one by one, which is rather time-consuming. Through combining different face detectors, we can reduce the processing time dramatically. The video shows that our method could detect the faces with different angles correctly and efficiently.

Back to Projects Page