342 A Wavelet-Transform Based Asker Identification System for Smart Multi-point Tele-conferences

Abstract

This paper proposes a wavelet-based asker identification system for smart multi-point tele-conferences. Our system is composed of three processors. The first processor is called an “ask” action detector, which is used to decide whether an “ask” action is occurring. If an “ask” action is detected, the input image is sent to the face detector that detects the human face from the complex background. The detected face is fed to a wavelet-based face identification processor to recognize the asker.