S-RL Toolbox: Environments, Datasets and Evaluation Metrics for State Representation Learning
This work solves the problem of inconsistent benchmarking for researchers in robotics and control, though it is incremental as it builds on existing methods without introducing new algorithms.
The paper addresses the lack of standardized evaluation in state representation learning by providing a toolbox with environments, datasets, and metrics to facilitate iterative learning and evaluation in reinforcement learning settings.
State representation learning aims at learning compact representations from raw observations in robotics and control applications. Approaches used for this objective are auto-encoders, learning forward models, inverse dynamics or learning using generic priors on the state characteristics. However, the diversity in applications and methods makes the field lack standard evaluation datasets, metrics and tasks. This paper provides a set of environments, data generators, robotic control tasks, metrics and tools to facilitate iterative state representation learning and evaluation in reinforcement learning settings.