[1805.07914] Imitating Latent Policies from Observation
Abstract:In this paper, we describe a novel approach to imitation learning that infers latent policies directly from state observations. We introduce a method that characterizes the causal effects of latent actions on observations while simultaneously predicting their likelihood. We then outline an action alignment procedure that leverages a small amount of environment interactions to determine a mapping between the latent and real-world actions. We show that this corrected labeling can be used for imitating the observed behavior, even though no expert actions are given. We evaluate our approach within classic control environments and a platform game and demonstrate that it performs better than standard approaches. Code for this work is available at this https URL.
View PDF HTML (experimental) Abstract:In this paper, we describe a novel approach to imitation learning that infers latent policies directly from state observations. We introduce a method that characterizes the causal effects of latent actions on observations while simultaneously predicting their likelihood. We then outline an action alignment procedure that leverages a small amount of environment interactions to determine a mapping between the latent and real-world actions. We show that this corrected labeling can be used for imitating the observed behavior, even though no expert actions…
saved by
related reading
- Behavioral Cloning from Observationarxiv.org
- 2312.10812arxiv.org
- Generative Adversarial Imitation Learningarxiv.org
- Learning to Imitate | SAIL Blogai.stanford.edu
- Ch. 21 - Imitation Learningunderactuated.mit.edu
- InfoGAIL: Interpretable Imitation Learning from Visual Demonstrationsarxiv.org
- State of Robot Learning, December 2025vedder.io
- Learning to Act without Actionsarxiv.org
- Generative Adversarial Imitation Learning.pdfcs.stanford.edu
- 1011.0686arxiv.org
- Reinforcement Learning via Implicit Imitation Guidancearxiv.org
- [2606.09758] Difference-Aware Retrieval Policies for Imitation Learningarxiv.org