ISSS608 2017-18 T3 Assign Sun Shuangtian Kasios’ Recording Investigation
VAST 2018 Mini Challenge 1
|
|
|
|
|
|
|
|
Contents
Geospatial Distribution of the Recordings
In our previous analysis, we could visualize the Rose-crested Blue Pipits in the Preserve appeared mostly in northeast of the Preserve. Whereas according to the recordings provided by Kasios, Blue Pipits appeared around the whole Preserve. This makes us doubt the authenticity of the Blue Pipit recordings provided by Kasios. Further analysis needs to be done to identify if the recordings provided by Kasios are really the sounds of the Blue Pipits.
Spectrogram of True Blue Pipit
In the collection of bird recordings from the Preserve which has accurate identifications, we randomly choose 4 Rose-crested Blue Pipit Song recordings and 4 Rose-crested Blue Pipit Call recordings. Then we use R package seewave to break the recordings and detect the frequencies respectively.
Song Frequency
Call Frequency
Pattern Summary
From the spectrograms above We could conclude that the frequencies of Blue Pipit songs and calls are both around 5KHz.
Spectrogram of Recordings from Kasios
Acceptable Recordings with Frequency around 5KHz (5 out of 15)
Fake Recordings with Frequency far from 5KHz (10 out of 15)
Frequency Comparison
Compare the frequency of the 15 recordings and the frequency of Blue Pipit song/call, we notice that for most of the 15 recordings provided by Kasios, the frequency ranges are not around 5KHz. This further proves that the recordings provided Kasios are not Rose-crested Blue Pipits.
