Abstract
We propose an integrated system using active learning for audio-to-score alignment. Audio-to-score alignment is a fundamental task in music information retrieval. Although various machine learning techniques have been applied to this task, it is not the case for active learning. To show how beneficial active learning is in audio-to-score alignment, we demonstrate a system that integrates it with dynamic time warping, a commonly used algorithm for time series alignment. We propose a simple parametric model for selecting queries-a crucial step in active learning. We evaluate the system using synthesized audio as well as real performances. The alignment accuracy is improved with a range from 20% to 50% using only less than 10% query instances, a promising result that hopefully can inspire the creation of a collaborative framework between human and machine for audio-to-score alignment in the future.
Original language | English (US) |
---|---|
Title of host publication | Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 796-799 |
Number of pages | 4 |
ISBN (Electronic) | 9781509061662 |
DOIs | |
State | Published - Jan 31 2017 |
Externally published | Yes |
Event | 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 - Anaheim, United States Duration: Dec 18 2016 → Dec 20 2016 |
Other
Other | 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016 |
---|---|
Country | United States |
City | Anaheim |
Period | 12/18/16 → 12/20/16 |
Fingerprint
Keywords
- Active learning
- Audio-to-score alignment
- Dynamic time warping
ASJC Scopus subject areas
- Artificial Intelligence
- Computer Networks and Communications
- Computer Science Applications
Cite this
An active learning approach to audio-to-score alignment using dynamic time warping. / Chuan, Ching-Hua.
Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016. Institute of Electrical and Electronics Engineers Inc., 2017. p. 796-799 7838246.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - An active learning approach to audio-to-score alignment using dynamic time warping
AU - Chuan, Ching-Hua
PY - 2017/1/31
Y1 - 2017/1/31
N2 - We propose an integrated system using active learning for audio-to-score alignment. Audio-to-score alignment is a fundamental task in music information retrieval. Although various machine learning techniques have been applied to this task, it is not the case for active learning. To show how beneficial active learning is in audio-to-score alignment, we demonstrate a system that integrates it with dynamic time warping, a commonly used algorithm for time series alignment. We propose a simple parametric model for selecting queries-a crucial step in active learning. We evaluate the system using synthesized audio as well as real performances. The alignment accuracy is improved with a range from 20% to 50% using only less than 10% query instances, a promising result that hopefully can inspire the creation of a collaborative framework between human and machine for audio-to-score alignment in the future.
AB - We propose an integrated system using active learning for audio-to-score alignment. Audio-to-score alignment is a fundamental task in music information retrieval. Although various machine learning techniques have been applied to this task, it is not the case for active learning. To show how beneficial active learning is in audio-to-score alignment, we demonstrate a system that integrates it with dynamic time warping, a commonly used algorithm for time series alignment. We propose a simple parametric model for selecting queries-a crucial step in active learning. We evaluate the system using synthesized audio as well as real performances. The alignment accuracy is improved with a range from 20% to 50% using only less than 10% query instances, a promising result that hopefully can inspire the creation of a collaborative framework between human and machine for audio-to-score alignment in the future.
KW - Active learning
KW - Audio-to-score alignment
KW - Dynamic time warping
UR - http://www.scopus.com/inward/record.url?scp=85015407514&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85015407514&partnerID=8YFLogxK
U2 - 10.1109/ICMLA.2016.2
DO - 10.1109/ICMLA.2016.2
M3 - Conference contribution
AN - SCOPUS:85015407514
SP - 796
EP - 799
BT - Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016
PB - Institute of Electrical and Electronics Engineers Inc.
ER -