Multimodal speaker localization from omnidirectional videos
Pascal Reuse, Mihai Gurban, Ivar Austvoll, Jean‐Philippe Thiran · Infoscience (Ecole Polytechnique Fédérale de Lausanne) · 2009
The use of omnidirectional cameras for videoconferencing promises to simplify the hardware setup necessary for large groups of participants. We investigate the use of a multimodal speaker detection algorithm on audio-visual sequences captured with such a camera, in particular, an algorithm that uses the audio energy together with the optical flow. We analyze several types of optical flow methods to determine the one which is appropriate to the omnidirectional context.