Panorama: a what I see is what I want contactless visual interface

In a face to face meeting, participants are free to chose who they look at and who they listen to. In a videoconference, participants can only see or hear the audio-visual signal that the distant participants have decided or have been able to transmit. We present Panorama, a contactless visual man machine interface which allows to explore a distant audio-visual scene. This scene, which corresponds to the end user interest, is complementary to the audio-visual scene obtained by automatically framing a person with LISTEN, the previous visual man machine interface we have developed. Panorama is an extension of LISTEN. The face detection and tracking techniques involved in LISTEN have been improved and adapted to meet the requirements of the exploration interface. Face detection has been extended to oriented face and tracking meets the needs of an exploration function convention.