SimWeaver: Zero-Shot RGB Sim-to-Real for Deformable Manipulation
It provides a practical solution for training manipulation policies on deformable objects without real-world data, addressing a key bottleneck in robotics.
SimWeaver achieves zero-shot RGB sim-to-real for deformable manipulation, reaching above 80% per-task and 91% average real-world success across 5 tasks using only 200 simulated demonstrations per task, without real-world fine-tuning.
RGB sim-to-real for deformable manipulation has remained largely unsolved without real-world fine-tuning. We present SimWeaver, which trains zero-shot RGB VLA policies on 200 simulated demonstrations per task, reaching above 80% per-task and 91% average real-world success across 5 diverse deformable tasks including plastic-bag manipulation, without teleoperation or per-task calibration. SimWeaver combines a reliable measurement-backed simulator (SimWeaver-Sim) with an extensible asset framework supporting single-image generation(SimWeaver-Asset), a deterministic topology-aware trajectory synthesizer (SimWeaver-Syn), and a sim-to-real protocol with ISP-aware photometric augmentation (SimWeaver-Real). On silk grasping, the sim-trained policy reaches 100% under visual distribution shifts where real-data baselines drop to 9-70%, at two orders of magnitude lower per-trajectory cost. We will release SimWeaver and a representative asset subset. Project page: https://simweaver.github.io/