Difference between revisions of "Acoustic Analysis"
| Line 14: | Line 14: | ||
5) Gives each of these peaks a time-stamped score based on amount of change in waveform relative to neighboring samples |
5) Gives each of these peaks a time-stamped score based on amount of change in waveform relative to neighboring samples |
||
| + | |||
In the spectragram: |
In the spectragram: |
||
| Line 23: | Line 24: | ||
8) Compares waveform candidates to spectrum candidates, keeps those where time scores align |
8) Compares waveform candidates to spectrum candidates, keeps those where time scores align |
||
| − | Calculates a burst score (burst strength) for remaining candidates: |
||
| − | + | Calculates a burst score (burst strength) for remaining candidates. Using a linear model trained on the burst locations in TIMIT: |
|
| + | 10) Selects candidate with highest burst score. |
||
| − | 10) |
||
| − | 10) If |
+ | 10) If there is no burst, returns a burst_time of -1. |
Revision as of 13:26, 16 April 2018
Burst detection (burst.py)
This script does the following things:
1) Defines a function burst() which takes as input three arguments: a soundfile, start time, and end time.
2) Resamples sound file to 16,000 Hz.
In the soundfile waveform:
3) Within the window specified by the start and end times, goes through each sample one by one and determines whether it is a peak or valley.
4) Finds three biggest valleys in the waveform (corresponding to pressure peaks) within specified time window.
5) Gives each of these peaks a time-stamped score based on amount of change in waveform relative to neighboring samples
In the spectragram:
6) Now, takes a Mel frequency spectrum with freq above 300 Hz, in 5 ms windows.
7) Compares this window to the next 5 ms spectral window, and selects top three candidates with most change in the spectrum.
8) Compares waveform candidates to spectrum candidates, keeps those where time scores align
Calculates a burst score (burst strength) for remaining candidates. Using a linear model trained on the burst locations in TIMIT:
10) Selects candidate with highest burst score.
10) If there is no burst, returns a burst_time of -1.
