Skip to main content

Table 2 The specific information of the database

From: Neural network-based non-intrusive speech quality assessment using attention pooling function

Conditions

Background noise

MNRU (Q = 5, 6, 10, 12, 15, 18, 24, 25, 30, 35, 36)

No background noise

MNRU (Q = 8, 14, 20, 26, 32)

Car noise: 15dB

G.729.1

No background noise

 

Office noise: 40dB

 

Babble noise: 40dB

 

Babble noise: 128dB

 

Car noise: 15dB

 

Street noise: 20dB

G.711, Codec1, Codec2

No background noise

 

Background music

 

Office noise

 

Babble noise

 

Interfering talker

 

Car noise

 

Office noise: 40dB

G.729, G.729E, Codec3, Codec4, Codec5, Codec6

No background noise

 

Car noise: 15dB