Chris@80
|
1
|
Chris@80
|
2 # Piper
|
Chris@80
|
3
|
Chris@80
|
4 ## A protocol for driving audio feature extractors
|
Chris@80
|
5
|
Chris@80
|
6 Piper is a protocol for audio analysis and feature extraction. That
|
Chris@80
|
7 is, the task of processing sampled audio data to produce descriptive
|
Chris@80
|
8 output (measurements or semantic observations).
|
Chris@80
|
9
|
Chris@80
|
10 Piper defines a data schema and API that can be used for remote audio
|
Chris@80
|
11 feature extraction services, or for feature extractors loaded directly
|
Chris@80
|
12 into a host application.
|
Chris@80
|
13
|
Chris@80
|
14 Piper is intended to be used
|
Chris@80
|
15
|
Chris@80
|
16 * as a programmatic interface for audio analysis and feature
|
Chris@80
|
17 extraction methods for web applications and servers
|
Chris@80
|
18
|
Chris@80
|
19 * to make [Vamp plugins](http://vamp-plugins.org), and feature
|
Chris@80
|
20 extractors written in other languages such as Javascript, available
|
Chris@80
|
21 through a service API or as loadable modules
|
Chris@80
|
22
|
Chris@80
|
23 The Piper schema is language- and serialisation-independent and the
|
Chris@80
|
24 API is transport-independent. We provide initial implementations using
|
Chris@80
|
25 JSON in Javascript and C++, and using Cap'n Proto in C++.
|
Chris@80
|
26
|
Chris@80
|
27 This repository contains the basic Piper schema. Implementations and
|
Chris@80
|
28 utilities can be found in related repositories.
|
Chris@80
|
29
|
Chris@80
|
30 ## Authors and licensing
|
Chris@80
|
31
|
Chris@80
|
32 Piper was made by Lucas Thompson and Chris Cannam at the Centre for
|
Chris@80
|
33 Digital Music, Queen Mary, University of London.
|
Chris@80
|
34
|
Chris@80
|
35 Copyright (c) 2015-2016 Queen Mary, University of London, provided
|
Chris@80
|
36 under a BSD-style licence. See the file COPYING for details.
|
Chris@80
|
37
|