NMF-weighted SRP for multi-speaker direction of arrival estimation: robustness to spatial aliasing while exploiting sparsity in the atom-time domain

EURASIP Journal on Audio, Speech, and Music Processing

Table 2 F-measure and mean azimuth error (MAE) for the supervised NMF-SRP (SNMF-SRP) for varying D_SNMF-SRP and β_SNMF-SRP

\(\frac {\beta _{\text {SNMF-SRP}}}{D_{\text {SNMF-SRP}}}\)	0	20	40
50	0.9/7.69	0.95/6.22	0.93/6.57
90	0.91/7.66	0.97/4.95	0.94/7.15
130	0.92/6.82	0.98/4.5	0.97/4.85