Hello everyone, welcome to Week 11 of my senior project. This week, I worked
on audio analysis, writing slides and presenting. Today I presented in the
Library of Congress for my internship, and I will present at school tomorrow.
After loading all the audio into the Sonic Visualizer, software that displays the frequency content of the audio, we start to analyze the audio by listening to the produced music and looking at the spectrum.
Audio Analysis: Comparison of Options
We observe the data of frequency in the ranges
– Green (Simply Average) has the most abrupt values at the high frequency range
– That means there are many outliers!
– Solutions to outlier numbers
– The ‘Outlier’ option successfully reduces the noise outliers at the high frequency range
– So does the “brightness value” option, since the brightness of the value are also taken into account
– Furthermore, the “brightness and outlier” option in the plugin further reduces the extreme values
– This implies that putting the features together can make them both more effective!
For the smoothing methods, we get different methods work differently on the
average option, and averaging smoothing method works better on the outlier
option. Besides, smoothing method proves to change more than the options.
Writing slides and paper
For the slides and my paper for the Library of Congress, I added the future
research and acknowledgement after writing the algorithm and analysis part. I
will keep exploring and contributing to weaver and image processing.
Today, I presented my slides to the people in the Library of Congress. The
whole process is smooth and the audience asked interesting questions such as
how was the debugging process. I really enjoyed the whole presentation.
During this internship, I explored new methods for better image processing. I want to thank my mentor Peter Alyea and my advisor Dr. Ryan Grove.
Thank you so much for reading all my blogs!