RoboTwin
A robot simulation benchmark for dual-arm collaborative manipulation. It emphasizes the evaluation of two-handed coordination and fine manipulation capabilities, using task success rate as the core metric.
Timeline (last 90 days)
REKI reduces the oscillation episode rate on the RoboTwin benchmark from 18.6% to 4.1%.
The single-frame memoryless policy has a RoboTwin oscillation rate of 18.6%, injecting multi-frame history into the VLM backbone gives 5.8%, and REKI gives 4.1%.
In simulation environments, RoboRSI achieves the highest success rates on benchmarks such as LIBERO, LIBERO-PRO, LIBERO-Plus, and RoboTwin, exceeding the strongest baseline by 2.7 to 11.0 percentage points.
RoboTwin is a simulation benchmark for dual-arm collaborative manipulation, emphasizing two-handed coordination and fine manipulation.
The paper validates on three platforms: the simulation benchmarks RoboTwin and LIBERO, as well as a real robot platform.
All Facts (5)
REKI reduces the oscillation episode rate on the RoboTwin benchmark from 18.6% to 4.1%.
50%UnverifiedThe single-frame memoryless policy has a RoboTwin oscillation rate of 18.6%, injecting multi-frame history into the VLM backbone gives 5.8%, and REKI gives 4.1%.
50%UnverifiedIn simulation environments, RoboRSI achieves the highest success rates on benchmarks such as LIBERO, LIBERO-PRO, LIBERO-Plus, and RoboTwin, exceeding the strongest baseline by 2.7 to 11.0 percentage points.
50%UnverifiedRoboTwin is a simulation benchmark for dual-arm collaborative manipulation, emphasizing two-handed coordination and fine manipulation.
50%UnverifiedThe paper validates on three platforms: the simulation benchmarks RoboTwin and LIBERO, as well as a real robot platform.
50%