Sequential Monte Carlo fusion of sound and vision for speaker tracking | IEEE Conference Publication | IEEE Xplore