发表机构
Kaedim（Kaedim）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究通过对比两种重建方案，证明提高环境重建质量（如度量尺度、物理参数和视觉保真度）能显著缩小仿真与真实机器人评估的差距，相关系数从0.51提升至0.90。
AI 中文摘要
仿真评估越来越多地与真实世界评估并行用于机器人策略，因为它更便宜且更易于重复；然而，其价值取决于其结果与真实机器人的匹配程度。我们测试了我们的重建流程，该流程结合了度量尺度化的物体几何、编写的物理参数和场景重建，是否相对于默认的开源方案减少了仿真与真实机器人评分之间的差异。我们构建了一个双臂机器人单元的两个仿真版本：一个是编写的重建，使用估计度量尺度下的物体几何、投影纹理、编写的物理和自有的场景splat；另一个是基线，称为默认重建，使用生成式单图像网格的开源方案、引擎默认物理和高斯splat场景。两个重建使用相同的物体照片和场景视频。两个策略各运行五个任务，产生十个任务-策略对，我们称之为单元；每个单元在每个重建中运行二十次，所有其他设置保持不变。两个重建均与相同的真实试验进行评分比较，由第三方评估者评分。十个仿真和真实单元均值之间的皮尔逊相关系数，对于编写的重建为r=0.90，对于默认重建为0.51。平均评分误差，编写的重建为6.97个百分点，默认重建为17.54个百分点，减少了10.56个百分点。这些结果表明，通过更高的视觉保真度、编写的物理和度量尺度提高环境重建质量，使仿真更忠实于真实世界，并缩小了仿真到真实的差距。我们发布了测试工具、每次试验的评分、每个报告运行的配置，以及两个重建的资产和场景。
英文摘要
Simulated evaluation is increasingly used alongside real-world evaluation of robot policies because it is cheaper and easier to repeat; however, its value depends on how closely its outcomes track the real robot's. We test whether our reconstruction pipeline, combining metrically scaled object geometry, authored physical parameters and scene reconstruction, reduces disagreement between simulated and real robot scores relative to a default open-source recipe. We constructed two simulated versions of one bimanual robot cell: an authored reconstruction, using object geometry at estimated metric scale, projected textures, authored physics and our own scene splat; and a baseline, referred to as the default reconstruction, using the open-source recipe of a generative single-image mesh, engine-default physics and a Gaussian-splat scene. Both reconstructions use the same object photographs and scene video. Two policies ran five tasks each, giving ten task-policy pairs, which we call cells; each cell was run twenty times in each reconstruction with all other settings held fixed. Both reconstructions were scored against the same real trials, graded by a third-party evaluator. Pearson correlation between the ten simulated and real cell means is r = 0.90 for the authored reconstruction and 0.51 for the default. Mean score error is 6.97 percentage points for the authored reconstruction and 17.54 for the default, a reduction of 10.56 percentage points. These results show that improving the quality of the environment reconstruction through higher visual fidelity, authored physics and metric scale makes the simulation more faithful to the real world and narrows the sim-to-real gap. We release the harness, the per-trial scores, every reported run's configuration, and the assets and scenes of both reconstructions.
Commentsv2: added link to the code and data repository. Code: https://github.com/Kaedim/yam-sim-harness-oss
