Advances in Multimedia / 2013 / Article / Fig 7

Research Article

Real-Time Audio-Visual Analysis for Multiperson Videoconferencing

Figure 7

The HMM used for each person to estimate voice activity from visual cues. The hidden, binary variable indicates if the person is speaking or not. The probability of is estimated recursively using the previous estimate and the current observation .
175745.fig.007

You are browsing a BETA version of Hindawi.com. Click here to switch back to the original design.