BPN Based Likelihood Ratio Score Fusion for Audio-Visual Speaker Identification in Response to Noise

Figure 12

Shape based facial feature and MFCC based audio features for score fusion technique.