REVIEW 8 cited by
The Ingredients of Real-World Robotic Reinforcement Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
The success of reinforcement learning for real world robotics has been, in many cases limited to instrumented laboratory scenarios, often requiring arduous human effort and oversight to enable continuous learning. In this work, we discuss the elements that are needed for a robotic learning system that can continually and autonomously improve with data collected in the real world. We propose a particular instantiation of such a system, using dexterous manipulation as our case study. Subsequently, we investigate a number of challenges that come up when learning without instrumentation. In such settings, learning must be feasible without manually designed resets, using only on-board perception, and without hand-engineered reward functions. We propose simple and scalable solutions to these challenges, and then demonstrate the efficacy of our proposed system on a set of dexterous robotic manipulation tasks, providing an in-depth analysis of the challenges associated with this learning paradigm. We demonstrate that our complete system can learn without any human intervention, acquiring a variety of vision-based skills with a real-world three-fingered hand. Results and videos can be found at https://sites.google.com/view/realworld-rl/
Forward citations
Cited by 8 Pith papers
-
Geometry of Neural Reinforcement Learning in Continuous State and Action Spaces
For wide two-layer linearized neural policies in deterministic continuous RL, the locally attainable states concentrate on a manifold of dimension at most 2da+1, independent of the state dimension.
-
From Prior to Pro: Efficient Skill Mastery via Distribution Contractive RL Finetuning
Residual off-policy RL with selective BC regularization and value-guided sampling contracts a pretrained generative robot policy around successful actions, reaching high success on hard long-horizon tasks from pixels ...
-
Auto-exploration for online reinforcement learning
New parameter-free SPMD algorithms achieve the first algorithm-independent O(ε⁻²) sample complexity for online discounted RL under a mixing-optimal-policy assumption.
-
SafeMimic: Towards Safe and Autonomous Human-to-Robot Imitation for Mobile Manipulation
SafeMimic enables a mobile robot to safely and autonomously adapt a single third-person human video into a successful multi-step manipulation strategy.
-
CARoL: Context-aware Adaptation for Robot Learning
CARoL measures task similarity by state-transition prediction errors and uses those similarities to weight prior policies, value functions, or actor-critic knowledge when adapting to a new task.
-
Flow-based Domain Randomization for Learning and Sequencing Robotic Skills
A normalizing-flow sampling distribution, trained with entropy-regularized reward maximization, improves domain coverage and sim-to-real transfer over Gaussian, beta, and interval-based learned domain randomization.
-
SimLauncher: Launching Sample-Efficient Real-world Robotic Reinforcement Learning via Simulation Pre-training
Simulation-pretrained policies, with digital-twin demos for critic bootstrapping and action proposals, cut real-world RL training time while reaching near-perfect success on three manipulation tasks.
-
Residual Reward Models for Preference-based Reinforcement Learning
Combining a hand-designed or learned prior reward with a preference-trained residual improves sample efficiency and final performance in preference-based reinforcement learning.
Discussion (0). Continue with ORCID to comment.