BriefGPT.xyz
Aug, 2022
一种基于概念的音乐分类器的“音乐学家友好型”解释技术
Concept-Based Techniques for "Musicologist-friendly" Explanations in a Deep Music Classifier
HTML
PDF
Francesco Foscarin, Katharina Hoedt, Verena Praher, Arthur Flexer, Gerhard Widmer
TL;DR
本研究提出了两种解释深度学习音乐数据的高级概念方法，包括有监督学习和无监督学习。对一个已有的符号作曲家分类系统进行演示并展示其局限性。
Abstract
Current approaches for explaining
deep learning
systems applied to
musical data
provide results in a low-level feature space, e.g., by highlighting potentially relevant time-frequency bins in a spectrogram or tim
→