This paper proposes a new music source separation (MSS) model based on an architecture
with MLP-Mixer that leverages multilayer perceptrons (MLPs). Most of the recent MSS …

Vocal music teaching through E-learning platforms and in-person classrooms requires
intense attention and practice. The vocals and strings with/without instrument support are …

This paper presents MDX-Mixer, which improves music demixing (MDX) performance by
leveraging source signals separated by multiple existing MDX models. Deep-learning …

Vocal music practices using the internet and mobile computing rely on the distinguishable
pitch, note, and musical library information for ease of understanding. For addressing this …

本稿では, 音楽音響信号と, そこから複数の既存の音楽音源分離 (MDX) モデルによって分離された
音源信号とを, 時間変化する重みで混合してより高い分離性能を達成するシステム MDX-Mixer …