Sequential Disentanglement by Extracting Static Information From A Single Sequence Element

Nimrod Berman, Ilan Naiman, Idan Arbiv, Gal Fadlon, Omri Azencot

Research output: Contribution to journalConference articlepeer-review

Abstract

One of the fundamental representation learning tasks is unsupervised sequential disentanglement, where latent codes of inputs are decomposed to a single static factor and a sequence of dynamic factors. To extract this latent information, existing methods condition the static and dynamic codes on the entire input sequence. Unfortunately, these models often suffer from information leakage, i.e., the dynamic vectors encode both static and dynamic information, or vice versa, leading to a non-disentangled representation. Attempts to alleviate this problem via reducing the dynamic dimension and auxiliary loss terms gain only partial success. Instead, we propose a novel and simple architecture that mitigates information leakage by offering a simple and effective subtraction inductive bias while conditioning on a single sample. Remarkably, the resulting variational framework is simpler in terms of required loss terms, hyper-parameters, and data augmentation. We evaluate our method on multiple data-modality benchmarks including general time series, video, and audio, and we show beyond state-of-the-art results on generation and prediction tasks in comparison to several strong baselines. Code is at GitHub.

Original languageAmerican English
Pages (from-to)3539-3564
Number of pages26
JournalProceedings of Machine Learning Research
Volume235
StatePublished - 1 Jan 2024
Event41st International Conference on Machine Learning, ICML 2024 - Vienna, Austria
Duration: 21 Jul 202427 Jul 2024

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Software
  • Control and Systems Engineering
  • Statistics and Probability

Cite this