Joint Detection and Classification of Singing Voice Melody Using Convolutional Recurrent Neural Networks
Singing melody extraction essentially involves two tasks: one is detecting the activity of a singing voice in polyphonic music, and the other is estimating the pitch of a singing voice in the detected voiced segments. In this paper, we present a joint detection and classification (JDC) network that...
Main Authors: | Sangeun Kum, Juhan Nam |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-03-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/9/7/1324 |
Similar Items
-
Vocal Melody Extraction via HRNet-Based Singing Voice Separation and Encoder-Decoder-Based F0 Estimation
by: Yongwei Gao, et al.
Published: (2021-01-01) -
Singing Voice Detection in Electronic Music with a Long-Term Recurrent Convolutional Network
by: Raymundo Romero-Arenas, et al.
Published: (2022-07-01) -
Research on Singing Voice Detection Based on a Long-Term Recurrent Convolutional Network with Vocal Separation and Temporal Smoothing
by: Xulong Zhang, et al.
Published: (2020-09-01) -
Singing Transcription from Polyphonic Music Using Melody Contour Filtering
by: Zhuang He, et al.
Published: (2021-06-01) -
Exploring Channel Properties to Improve Singing Voice Detection with Convolutional Neural Networks
by: Wenming Gui, et al.
Published: (2021-12-01)