DID YOU KNOW? DTIC has over 3.5 million final reports on DoD funded research, development, test, and evaluation activities available to our registered users. Click HERE
to register or log in.
Head Related Transfer Function Approximation Using Neural Networks.
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Pagination or Media Count:
This thesis determines whether an artificial neural network ANN can approximate the Armstrong Aerospace Medical Research Laboratories AAMRL head related transfer functions HRTF data obtained from research at AAMRL during the fall of 1988. The first test determines whether HRTF lends any support in sound localization when compared to no HRTF Interaural Time Delay only. There is a statistically significant interaction between the location of the sound and whether the HRTF or no HRTF is used. When this interaction is removed using the alternate F.Value, the statistics give the result of equal means for the filters and azimuth. This means that at certain angles of azimuth, the HRTF either provides no advantage at all or hinders localization capabilities. Adding in the corrections for reversals changes the results to where the means of the azimuth are not statistically equal. The reversal corrections will inherently reduce the error results. However, comparing the number of reversals indicates an advantage of using the HRTF over no HRTF. The second test determines whether HRTF and ANN lend the same amount of information in sound localization when compared to each other. With reversal corrections included, a statistical advantage is not indicated, and the means of the two filters are statistically equal. Also there is a statistically significant interaction between the location of the sound and whether the AAMRL HRTF or ANN HRTF is used. Comparing the number of reversals does not indicate a large difference between the AAMRL HRTF and the RBF HRTF.
APPROVED FOR PUBLIC RELEASE