深度卷积网络中基于身份的模式:生成对抗音系学和重叠
Identity-Based Patterns in Deep Convolutional Networks: Generative Adversarial Phonology and Reduplication
摘要:
本文使用深度卷积神经网络对语音中基于身份的模式(或复制)的无监督学习进行建模,该模式被称为从原始连续数据中的重复。我们使用了ciwGAN架构Beguplant(2021 a; arXiv:2006.02951),其中语音中有意义表示的学习来自CNN生成信息数据的要求。我们提出了一种对语音训练的CNN进行wug-test的技术,并基于四个生成测试,认为该网络学习在其潜在空间中表示基于身份的模式。通过在潜在空间中仅操作两个分类变量,我们可以主动地将未重复的形式转换为重复的形式,而在大多数情况下不会对输出进行其他实质性更改。我们还认为,网络扩展了基于身份的模式,以未观察到的数据。探索基于身份的模式如何在CNN中出现有意义的表示,以及训练范围之外的潜在空间变量如何与输出中的基于身份的模式相关,对神经网络的可解释性具有普遍意义。
英文摘要:
This paper models unsupervised learning of an identity-based pattern (or copying) in speech called reduplication from raw continuous data with deep convolutional neural networks. We use the ciwGAN architecture Beguš (2021a; arXiv:2006.02951) in which learning of meaningful representations in speech emerges from a requirement that the CNNs generate informative data. We propose a technique to wug-test CNNs trained on speech and, based on four generative tests, argue that the network learns to represent an identity-based pattern in its latent space. By manipulating only two categorical variables in the latent space, we can actively turn an unreduplicated form into a reduplicated form with no other substantial changes to the output in the majority of cases. We also argue that the network extends the identity-based pattern to unobserved data. Exploration of how meaningful representations of identity-based patterns emerge in CNNs and how the latent space variables outside of the training range correlate with identity-based patterns in the output has general implications for neural network interpretability.
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。