Linear Disentangled Representations and Unsupervised Action Estimation
arXiv:2008.07922
Abstract
Disentangled representation learning has seen a surge in interest over recent times, generally focusing on new models which optimise one of many disparate disentanglement metrics. Symmetry Based Disentangled Representation learning introduced a robust mathematical framework that defined precisely what is meant by a "linear disentangled representation". This framework determined that such representations would depend on a particular decomposition of the symmetry group acting on the data, showing that actions would manifest through irreducible group representations acting on independent representational subspaces. Caselles-Dupre et al [2019] subsequently proposed the first model to induce and demonstrate a linear disentangled representation in a VAE model. In this work we empirically show that linear disentangled representations are not generally present in standard VAE models and that they instead require altering the loss landscape to induce them. We proceed to show that such representations are a desirable property with regard to classical disentanglement metrics. Finally we propose a method to induce irreducible representations which forgoes the need for labelled action sequences, as was required by prior work. We explore a number of properties of this method, including the ability to learn from action sequences without knowledge of intermediate states and robustness under visual noise. We also demonstrate that it can successfully learn 4 independent symmetries directly from pixels.
References in corpus (13)
- Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
- Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations
- Disentangling by Factorising
- Towards a Definition of Disentangled Representations
- Learning Deep Disentangled Embeddings with the F-Statistic Loss
- Contingency-Aware Exploration in Reinforcement Learning
- Imitating Latent Policies from Observation
- Independently Controllable Factors
- Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
- Learning what you can do before doing anything
- Flatland: a Lightweight First-Person 2-D Environment for Reinforcement Learning
- Learning Group Structure and Disentangled Representations of Dynamical Environments
- Disentangling by Subspace Diffusion