Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

 
Session Overview
Session
#WkshpAM2: Introduction to Sound Analysis Workshop
Time:
Tuesday, 23/Jul/2019:
9:00am - 12:00pm

Location: Duquesne University, Gumberg Library, Room 408
600 Forbes Ave, Pittsburgh, PA 15282

Show help for 'Increase or decrease the abstract text size'
Presentations

Introduction to Sound Analysis

Tanya E. Clement1, Brian McFee2

1University of Texas at Austin, United States of America; 2New York University, Steinhardt

Learning to work with computational approaches to sound studies must begin with a basic understanding about the ways in which sound is represented as data in the computational environment. The sound analysis workflows that we will share with participants will include basic steps for employing common, free and open-source, python libraries for accessing, processing, analyzing, and visualizing audio data. These workflows will be presented as Jupyter notebooks that users can easily adapt and run from local computers during and after the workshop. Topics that we will cover in this 3-hour workshop include:

  1. An introduction to setting up your computer for sound analysis with a brief introduction to using Binder and the Docker image we will set up for participants;
  2. An introduction to sound analysis using python, python libraries (such as McFee’s Librosa python package https://github.com/librosa/librosa/), and Jupyter notebooks;
  3. A demonstration with an example slicing and dicing audio that will introduce participants to basic audio concepts in signal processing such as sampling rates and the fundamental frequency;
  4. A hands-on group activity with a data set we have constructed including poetry performances from the SpokenWeb project (https://spokenweb.ca/) that will include:
  1. Silence/non-silence detection and auto-segmentation using unsupervised learning approaches such as K-means;
  2. Vocal / non-vocal detection using a pre-designed model and data set, during which participants will be introduced to how a supervised model is trained;
  3. Speaker segmentation and grouping using a pre-designed model and data set, during which participants will be introduced to how a supervised model is trained;

This workshop is appropriate for beginners to programming and sound analysis. No experience needed. Participants should bring their own laptops unless a computer lab can be provided.



 
Contact and Legal Notice · Contact Address:
Privacy Statement · Conference: ACH 2019
Conference Software - ConfTool Pro 2.6.128+TC
© 2001 - 2019 by Dr. H. Weinreich, Hamburg, Germany