Skip to content
For Sustainable Design Studio fall 2017
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
0109 one_hkatu_recording
Hämeenkatu manually classified data.txt
LICENSE
README.md
calibrate_dbs.m
confusion_matrix.png
dbA_avg.m
dba_test1_sound2.wav
dba_test1_sound3.wav
dba_test1_sound4.wav
dba_test1_sound5.wav
dba_test1_sound6.wav
filterA.m
gender_recognition.py
license.txt
plot_dbs.m
salamon-cnn.h5
us8k_extract.py
us8k_salamon.py
voice.csv
voicemodel.h5

README.md

For Sustainable Design Studio / Urban Informatics course Fall 2017

Building a CNN to classify the UrbanSounds8K dataset:

Calculating dBA weighted SPL for a calibrated microphone:

  • filterA.m: Matlab A-weighting filter by M.Sc. Eng. Hristo Zhivomirov
  • license.txt: License for filterA.m
  • dbA_avg.m: calculates dBA SPL of a .wav file
  • calibrate_dbs.m: calculates calibration value for accurate SPL measurements based on recordings where a calibration device was used to read wanted SPL was present next to the microphone
  • plot_dbs.m: plots dBA SPL values of Hämeenkatu recording excerpts
  • 0109one_hkatu_recording: manually annotated data from Hämeenkatu on one Saturday morning, next to Stockmann. Move contents of this folder to same folder as plot_dbs.m to use
  • dbatest1_sound1.wav...dbatest2_sound6.wav: calibration recordings for calibrate_dbs.m

Gender recognition from a csv file with voice frequency-space data:

Manual annotation of Hämeenkatu data:

  • Hämeenkatu manually classified data.txt: manually classified categories of all Hämeenkatu data
You can’t perform that action at this time.