Comparing MFCC with AR features

Now we compare MFCC feature chains with an AR feature chain using data from both AR and MFCC-based circularly-stationary models. In Figure 10.3, we show the acid test results for data generated according to the circular AR model. The accuracy of the AR feature chain is far superior.
Figure 10.3: Acid test performance of AR features (left) and MFCC features (right) for data generated according to the circular AR model.
\includegraphics[height=3.0in,width=3.0in]{ar_ar_mftest.eps} \includegraphics[height=3.0in,width=3.0in]{mf_ar_mftest.eps}
In Figure 10.4, we show the acid test results for data generated according to the circular MFCC model. The accuracy of the MFCC feature chain is far superior.
Figure 10.4: Acid test performance of AR features (left) and MFCC features (right) for data generated according to the circular AR model.
\includegraphics[height=3.0in,width=3.0in]{ar_mf_mftest.eps} \includegraphics[height=3.0in,width=3.0in]{mf_mf_mftest.eps}
These experiments underscore the importantce of matching the feature extraction to the data. Notice also that by using the ``wrong" freatures, the result is PDF estimation accuracy. For the mis-matched feature, log-likelihood for the projected PDF is always biased significantly lower, never higher! This is because the true PDF is the likelihood function with the highest average likelihood.

Baggenstoss 2017-05-19