Proceedings of the 7th Conference on Sound and Music Technology (CSMT)

The book presents selected papers that have been accepted at the seventh Conference on Sound and Music Technology (CSMT) in December 2019, held in Harbin, Hei Long Jiang, China. CSMT is a domestic conference focusing on audio processing and understanding with bias on music and acoustic signals. The primary aim of the conference is to promote the collaboration between art society and technical society in China. The organisers of CSMT hope the conference can serve as a platform for interdisciplinary research. In this proceeding, the paper included covers a wide range topic from speech, signal processing and music understanding, which demonstrates the target of CSMT merging arts and science research together.

Inhalt
Bandwidth Extension WaveNet for Bone-Conducted Speech Enhancement.- Naturalness evaluation of synthetic speech based on residual learning networks.- Detection of Operation Type and Order for Digital Speech.- Singing Voice Detection Using Multi-Feature Deep Fusion with CNN.- A Multi-task Learning Approach for MelodyExtraction.- A post-processing of onset detection based on verication with neural network.- Transfer learning for music classification and regression tasks using artist tags.