StereoEngine：无需立体相机即可扩展立体双机械臂数据
StereoEngine: Scaling Stereo Bimanual Robot Data without Stereo Cameras
AI总结：
针对机器人操纵领域大规模真实立体数据集稀缺的瓶颈，提出StereoEngine，通过两种互补方法合成立体数据，其生成数据可提升双机械臂操纵性能
摘要：
立体视觉正越来越多地应用于机器人操纵领域，越来越多的研究将其融入视觉-语言-动作（VLA）模型中。然而，大规模真实世界立体数据集的稀缺仍是扩展此类策略的根本瓶颈。近期的单目转立体生成方法主要针对VR内容、自动驾驶等场景，尚无研究生成适用于机器人操纵的逼真合成立体数据，目前立体监督主要来自仿真环境。我们提出StereoEngine，引入两种互补方法合成机器人操纵所需的立体数据：第一种方法利用估计的深度将单目（左）视频扭曲至右视图，并通过视频扩散模型修复遮挡区域，无需在真实立体数据上训练即可将单目演示转换为立体对；第二种方法在仿真环境中渲染立体对，联合生成两个逼真视图，无需任何真实视频作为输入即可生成左右一致的数据。StereoEngine在生成的立体质量上优于现有基线，将其合成数据用于策略中期训练或任务特定微调可显著提升下游双机械臂操纵性能。项目网站：this https URL
英文摘要：
Stereo vision is increasingly used for robotic manipulation, with a growing body of work incorporating it into Vision-Language-Action (VLA) models. However, the scarcity of large-scale, real-world stereo datasets remains a fundamental bottleneck for scaling such policies. Recent generative monocular-to-stereo methods target settings such as VR content and autonomous driving. No prior work produces realistic synthetic stereo data for robotic manipulation, where stereo supervision has come predominantly from simulation. We propose StereoEngine, which introduces two complementary methods to synthesize stereo data for robotic manipulation. In the first, we warp a monocular (left) video into the right view using estimated depth and inpaint the disoccluded regions with a video diffusion model, lifting monocular demonstrations into stereo pairs without training on real stereo. In the second, we render a stereo pair in simulation and jointly generate both photorealistic views, producing consistent left-right data without any real video as input. StereoEngine outperforms existing baselines in generated stereo quality, and using its synthetic data for policy mid-training or task-specific fine-tuning substantially improves downstream bimanual manipulation. Project website: https://stereoengine.github.io
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
