/Urban-Sound-Classification

Urban Sound Classification: With Random Forest, SVM, DNN, RNN, and CNN Classifiers

Primary LanguageJupyter Notebook

Paper

Authors Chih-Wei Chang and Benjamin Doran

Urban Sound Classification: With Random Forest, SVM, DNN, RNN, and CNN Classifiers

Presentation Slides

Urban Sound Classification: Comparision of Feature Extraction Techniques

Check Analysis

Download UrbanSound8K dataset from: https://serv.cusp.nyu.edu/projects/urbansounddataset/urbansound8k.html
(it is free, with thanks to Justin Salamon, Christopher Jacoby, and Juan Pablo Bello for creating the UrbanSound8K dataset)

Place tarfile in feature extractions directory. Run each feature extraction notebook. (May take a few hours)

Move resulting pickle files, ending in .p, to Models folder. Run desired model notebooks. (Time varies)

Move dataset_acc.p file to from model folder to plotting folder and run notebooks to generate dataset comparisons and training curve plot.