Differential DSP: An audio toolbox for end-to-end ml

The short-time Fourier transform (STFT) has been a staple of signal processing, often being the first step for many audio tasks. A very familiar process when using the STFT is the search for the best STFT parameters, as they often have significant side effects if chosen poorly. These parameters are often de ned in terms of an integer number of samples, which makes their optimization non-trivial. We present a toolbox that allows us to obtain gradients for commonly used audio filter parameters, and for STFT parameters with respect to arbitrary cost functions, thus enabling gradient descent optimization of quantities like the STFT window length or the STFT hop size. We do so for parameter values that stay constant throughout an input, but also for cases where these parameters have to dynamically change over time to accommodate varying signal characteristics.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Electrical & Computer Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2021

Author dc:creator

• Zhao, An

Contributors dc:contributor

• Smaragdis, Paris

• Classification
• Differential
• DSP
• DDSP
• Machine learning
• Audio

Statement dc:rights

• Copyright 2020 An Zhao

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/109453

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/109453