Robust Sounds of Activities of Daily Living Classification in Two-Channel Audio-Based Telemonitoring
Figure 6
Convergence behaviour of the SSA algorithm: differences between the optimum SNR and evolved SNR (a) and optimum value minus evolved (b), for various example sounds of daily life (as indicated) with background radio noise at pleasant listening level, using microphones 1 and 3. Two SDLs from each sound class were chosen at random.
