Dropout's Dream Land: Generalization from Learned Simulators to Reality

Zac Wellmer, James Kwok

Keywords: Reinforcement Learning

Abstract: A World Model is a generative model used to simulate an environment. World Models have proven capable of learning spatial and temporal representations of Reinforcement Learning environments. In some cases, a World Model offers an agent the opportunity to learn entirely inside of its own dream environment. In this work we explore improving the generalization capabilities from dream environments to reality (Dream2Real). We present a general approach to improve a controller's ability to transfer from a neural network dream environment to reality at little additional cost. These improvements are gained by drawing on inspiration from domain randomization, where the basic idea is to randomize as much of a simulator as possible without fundamentally changing the task at hand. Generally, domain randomization assumes access to a pre-built simulator with configurable parameters but oftentimes this is not available. By training the World Model using dropout, the dream environment is capable of creating a nearly infinite number of \textit{different} dream environments. Our experimental results show that Dropout's Dream Land is an effective technique to bridge the reality gap between dream environments and reality. Furthermore, we additionally perform an extensive set of ablation studies.

One-sentence Summary: A simple technique to bridge the sim2real gap between learned artificial neural network simulators and reality

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

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2109.08342/code)

Reviewed Version (pdf): https://openreview.net/references/pdf?id=1Go0EJYQyb