. Author manuscript; available in PMC: 2021 Jul 13.

Published in final edited form as: IEEE/ACM Trans Audio Speech Lang Process. 2019 Apr 29;27(7):1179–1188. doi: 10.1109/taslp.2019.2913512

TABLE III.

Performance Comparison Between the Baseline DNN and the Proposed Framework for Noise-Independent Models Trained on the IEEE Dataset. The Scores Denote the Percent Improvement in STOI Over Mixture

	Babble		Cafeteria
	DNN	AECNN-SM1	DNN	AECNN-SM1
5 dB	12	13.6	13.3	14.4
0 dB	17.1	21.1	18.1	21.1
−2 dB	18	23.5	18.7	22.5
5 dB	16.6	23.3	17.5	21.5