• Categories
• Main
• Core functions
• Common functions
• Data
• Demos
• Experiments
• Models
• Model stages
• Plot
• Signals
• FrAMBI
• Auxdata
• Cache

DEMO_JOERGENSEN2013
Demo for the multi-resolution speech-based envelope spectrum model

Description:

demo_joergensen2013 computes the signal-to-noise envelope-power ratio (SNRenv) for a speech sentence in noise at an SNR of -3 dB. The SNRenv can be then used to predict speech intelligibility.

Pcorrect indicates the probability of correctly understanding the sentence based on results from Joergensen et al. (2013).

This code produces the following output:

SNRenvs_noIOparameters =

   91.4732


SNRenvs_withIOparameters =

   91.4732


Pcorrect =

   99.8131

References:

S. Joergensen and T. Dau. Predicting speech intelligibility based on the signal-to-noise envelope power ratio after modulation-frequency selective processing. The Journal of the Acoustical Society of America, 130(3):1475--1487, 2011.

S. Jørgensen, S. D. Ewert, and T. Dau. A multi-resolution envelope power based model for speech intelligibility. J. Acoust. Soc. Am., 134(1):436--446, 2013.