Back To Schedule
Wednesday, November 20 • 15:00 - 15:25
Real-time dataflow for audio

Log in to save this to your schedule, view media, leave feedback and see who's attending!

A dataflow software architecture models computation as a directed graph, where the nodes are pure functions, and the edges between nodes are data. In addition to recent uses in deep learning, big data, and reactive programming, dataflow has long been an ideal fit for Digital Signal Processing (DSP). In a sense, deep learning's artificial neural networks can be thought of as DSP with large adaptive filters and non-linearities
Despite the success of dataflow in machine learning (ML) and DSP, there has not yet been to our knowledge a lightweight dataflow library that fulfills these requirements: small (under 50 Kbytes code), portable with few dependencies, open source, and most important: predictable performance suitable for embedded systems with real-time processing on the order of one millisecond per graph evaluation.
We describe a real-time dataflow architecture and initial C++ implementation for audio that meet these requirements, then explore the benefits of a unified view of ML and DSP. We also compare this C++ library approach to alternatives such as ROLI SOUL, which is based on a domain-specific programming language.

avatar for Domingo Hui

Domingo Hui

Intern, Google
I am a fourth year student studying Mathematics and Computer Science at the University of Waterloo. Interested in functional programming, low-level systems and real-time performance. I recently completed an internship implementing parts of the Android Audio framework using a dataflow... Read More →
avatar for Glenn Kasten

Glenn Kasten

software engineer, Google
Glenn Kasten is a software engineer in the Android media team, with a focus on low-­level audio and performance. His background includes real-­time operating systems and embedded applications, and he enjoys playing piano.

Wednesday November 20, 2019 15:00 - 15:25 GMT
Auditorium Puddle Dock, London EC4V 3DB