CITATION — REFERENCE ENTRY
Temporal convolutional networks for musical audio beat tracking — 2019 27th European Signal Processing Conference (EUSIPCO)
- Key
- davies2019-tcn-beat-tracking
- Authors
- Davies, Matthew E. P.; Böck, Sebastian
- Issued
- 2019
- Type
- paper-conference
- Container
- 2019 27th European Signal Processing Conference (EUSIPCO)
- Pages
- 1-5
Raw CSL JSON
{
"DOI": "10.23919/EUSIPCO.2019.8902578",
"page": "1-5",
"type": "paper-conference",
"title": "Temporal convolutional networks for musical audio beat tracking",
"author": [
{
"given": "Matthew E. P.",
"family": "Davies"
},
{
"given": "Sebastian",
"family": "Böck"
}
],
"issued": {
"date-parts": [
[
2019
]
]
},
"container-title": "2019 27th European Signal Processing Conference (EUSIPCO)"
}
Claims
-
Temporal convolutional networks achieve state-of-the-art beat tracking on many datasets, parallelise well for efficient training on large data, and need few weights, compared with the bidirectional LSTM approach.
"By contrasting our convolutional approach with the current state-of-the-art recurrent approach using Bidirectional Long Short-Term Memory, we demonstrate three highly promising attributes of TCNs for music analysis, namely: i) they achieve state-of-the-art performance on a wide range of existing beat tracking datasets, ii) they are well suited to parallelisation and thus can be trained efficiently even on very large training data; and iii) they require a small number of weights."
Available in