Representation Learning for Sequence Data with Deep Autoencoding Predictive Components

Junwen Bai, Weiran Wang, Yingbo Zhou, Caiming Xiong

Keywords: Mutual Information, Unsupervised Learning, Sequence Data, Masked Reconstruction

Abstract: We propose Deep Autoencoding Predictive Components (DAPC) -- a self-supervised representation learning method for sequence data, based on the intuition that useful representations of sequence data should exhibit a simple structure in the latent space. We encourage this latent structure by maximizing an estimate of \emph{predictive information} of latent feature sequences, which is the mutual information between the past and future windows at each time step. In contrast to the mutual information lower bound commonly used by contrastive learning, the estimate of predictive information we adopt is exact under a Gaussian assumption. Additionally, it can be computed without negative sampling. To reduce the degeneracy of the latent space extracted by powerful encoders and keep useful information from the inputs, we regularize predictive information learning with a challenging masked reconstruction loss. We demonstrate that our method recovers the latent space of noisy dynamical systems, extracts predictive features for forecasting tasks, and improves automatic speech recognition when used to pretrain the encoder on large amounts of unlabeled data.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Code: [ JunwenBai/DAPC](https://github.com/JunwenBai/DAPC) + [ 1 community implementation](https://paperswithcode.com/paper/?openreview=Naqw7EHIfrv)

Data: [LibriSpeech](https://paperswithcode.com/dataset/librispeech)