用大规模合成数据训练抓取模型，实现高精度真实场景抓取。
DexGraspNet 2.0: Learning Generative Dexterous Grasping in Large-scale Synthetic Cluttered Scenes
- 基于局部几何信息的扩散模型生成抓取姿态。
- 模拟实验中成功率超越所有基线方法，真实场景达90.7%。
- 零样本跨域迁移，适合机器人灵巧手抓取研究者。
由于数据稀缺，灵巧手在杂乱场景中的抓取仍具挑战性。为此，我们构建了一个大规模合成基准，包含1319个物体、8270个场景和4.27亿次抓取。除基准外，还提出一种两阶段抓取方法，利用条件于局部几何的扩散模型高效学习数据。所提生成式方法在仿真实验中优于所有基线。此外，借助测试时深度恢复，该方法实现了零样本模拟到现实的迁移，在真实杂乱场景中达到90.7%的灵巧抓取成功率。
原文摘要 · Abstract (English)
Grasping in cluttered scenes remains highly challenging for dexterous hands due to the scarcity of data. To address this problem, we present a large-scale synthetic benchmark, encompassing 1319 objects, 8270 scenes, and 427 million grasps. Beyond benchmarking, we also propose a novel two-stage grasping method that learns efficiently from data by using a diffusion model that conditions on local geometry. Our proposed generative method outperforms all baselines in simulation experiments. Furthermore, with the aid of test-time-depth restoration, our method demonstrates zero-shot sim-to-real transfer, attaining 90.7% real-world dexterous grasping success rate in cluttered scenes.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。