BigBIRD

该数据集包含了100多个在训练过程中未见过的全新物体，旨在用于评估在真实世界中抓取任务中对新物体的泛化能力。

This dataset contains over 100 completely novel objects that were unseen during the training phase, and is purpose-built to evaluate the generalization capability for novel objects in real-world grasping tasks.

搜集汇总

背景与挑战

背景概述

BigBIRD是一个大规模3D对象实例数据库，包含125个对象，每个对象提供600张高分辨率图像、RGB-D点云、姿态信息和分割掩码等多模态数据。该数据集旨在支持计算机视觉和机器人领域的3D重建、对象识别等研究，并计划持续扩展以促进方法开发。