EURASIP Journal on Audio, Speech, and Music Processing

Table 4 Comparsion of low-activity speaker counting performance under different acoustical conditions in terms of F1 score

From: Learning-based robust speaker counting and separation with the aid of spatial coherence

T60 (ms)	360			610
SNR (dB)	30	20	10	30	20	10	Avg.
Baseline 1	91.34	85.91	54.31	92.50	84.92	50.73	77.22
Baseline 2	97.65	89.49	64.27	96.29	86.47	65.78	82.73
Proposal 1	99.70	95.58	71.17	99.16	94.41	74.47	89.08
Proposal 2	99.70	98.43	78.21	99.75	98.10	80.22	92.40

Back to article page