EURASIP Journal on Audio, Speech, and Music Processing

Table 2 The SELD performance of the proposed method and comparison models for MIC dataset

From: Dual-branch attention module-based network with parameter sharing for joint sound event detection and localization

Models	ER_20°↓	F_20°↑	LE_CD↓	LR_CD↑	SELD↓
DCASE2020task3	0.690	41.3%	23.1°	62.4%	0.445
SELDnet	0.780	31.4%	27.3°	59.0%	0.503
Two-stage network	0.518	59.5%	14.2^°	69.6%	0.371
Cao_Surrey	0.471	61.5%	16.7°	75.4%	0.298
Nguyen_NTU	0.360	71.4%	12.1°	82.0%	0.223
Proposed method	0.364	71.9%	11.5°	80.3%	0.226

Back to article page