[outsig, fc] = jepsen2008preproc(insig,fs); [outsig, fc] = jepsen2008preproc(insig,fs,...);
insig | input acoustic signal. |
fs | sampling rate. |
Warning: This code cannot be verified. It has not been possible to tell from the desciption in the original paper nor from personal communication with the original authors what the correct parameter set used for the model is. This code is kept here as a reminder of the structure of the model, and may reappear in a future work if a verified parameter set can be established. The status of this piece of code is "not even wrong": http://en.wikipedia.org/wiki/Not_even_wrong.
jepsen2008preproc(insig,fs) computes the internal representation of the signal insig sampled with a frequency of fs Hz as described in Jepsen, Ewert and Dau (2008).
[outsig,fc]=jepsen2008(...) additionally returns the center frequencies of the filter bank.
The full Jepsen et al. (2008) model consists of the following stages:
- a heaphone filter to simulate the effect of a standard set of headphones.
- a middle ear filter to simulate the effect of the middle ear, and to convert to stapes movement.
- DRNL - Dual resonance non-linear filterbank.
- an envelope extraction stage done by half-wave rectification followed by low-pass filtering to 1000 Hz.
- an expansion stage
- an adaptation stage modelling nerve adaptation by a cascade of 5 loops.
- a modulation filterbank.
Any of the optinal parameters for drnl, ihcenvelope and adaptloop may be optionally specified for this function. They will be passed to the corresponding functions.
M. Jepsen, S. Ewert, and T. Dau. A computational model of human auditory signal processing and perception. J. Acoust. Soc. Am., 124(1):422-438, 2008.