arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1802.00680cs.LGcs.SDeess.ASstat.ML
收藏

一种基于势能模型的自然声生成模型

A Generative Model for Natural Sounds Based on Latent Force Modelling

Queen Mary University of London(伦敦玛丽女王大学)

William J. Wilkinson, Joshua D. Reiss, Dan Stowell

摘要:

自然声音的子带振幅包络分析的最新进展已经导致令人信服的合成,显示子带振幅是感知的重要组成部分。概率潜变量分析特别具有启发性,但现有的方法不包含有关振幅包络的物理行为的先验知识,例如指数衰减和反馈。我们使用潜在的力量建模,概率学习范式,将物理知识纳入高斯过程回归,跨频谱子带包络模型的相关性。我们通过明确地对多个时间步长的相关性进行建模来增强标准的潜在力模型方法。对这种先验知识的阐述加强了对潜在函数作为产生信号的源的解释。我们通过一个实验来研究这种解释,该实验表明,从我们的概率模型中采样产生的声音被认为比基于非负矩阵分解的类似模型产生的声音更真实,即使在我们的模型从重建误差的角度来看表现出色的情况下。

英文摘要:

Recent advances in analysis of subband amplitude envelopes of natural sounds have resulted in convincing synthesis, showing subband amplitudes to be a crucial component of perception. Probabilistic latent variable analysis is particularly revealing, but existing approaches don't incorporate prior knowledge about the physical behaviour of amplitude envelopes, such as exponential decay and feedback. We use latent force modelling, a probabilistic learning paradigm that incorporates physical knowledge into Gaussian process regression, to model correlation across spectral subband envelopes. We augment the standard latent force model approach by explicitly modelling correlations over multiple time steps. Incorporating this prior knowledge strengthens the interpretation of the latent functions as the source that generated the signal. We examine this interpretation via an experiment which shows that sounds generated by sampling from our probabilistic model are perceived to be more realistic than those generated by similar models based on nonnegative matrix factorisation, even in cases where our model is outperformed from a reconstruction error perspective.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑