Real2Sim2Real
A robot learning transfer pipeline that first calibrates or constructs a simulation environment using real data, then trains policies in simulation and deploys them back to the real robot, thereby bridging the sim-to-real gap.
Timeline (last 90 days)
Real2Sim2Real generally refers to a closed-loop pipeline in which real data is first used to calibrate or build a simulation environment (Real2Sim), and policies are then trained in simulation and deployed back to the real world (Sim2Real).
When combined with the Real2Sim2Real pipeline, AtomicRL improves real-world task performance by an average of 21.3%.
AtomicRL's average 21.3% real-world improvement depends on the effectiveness of the Real2Sim2Real transfer pipeline, whereas the improvements on simulation benchmarks mainly come from SAPO and BATE's improvements to credit assignment and exploration, so the two have different sources of contribution.
All Facts (3)
Real2Sim2Real generally refers to a closed-loop pipeline in which real data is first used to calibrate or build a simulation environment (Real2Sim), and policies are then trained in simulation and deployed back to the real world (Sim2Real).
50%UnverifiedWhen combined with the Real2Sim2Real pipeline, AtomicRL improves real-world task performance by an average of 21.3%.
50%UnverifiedAtomicRL's average 21.3% real-world improvement depends on the effectiveness of the Real2Sim2Real transfer pipeline, whereas the improvements on simulation benchmarks mainly come from SAPO and BATE's improvements to credit assignment and exploration, so the two have different sources of contribution.
50%