NVIDIA Isaac Sim Assets

该数据集包含NVIDIA Isaac Sim平台中用于机器人仿真与人工智能训练的数字资产资源，涵盖机器人模型、传感器、环境场景及物理属性四大类数据。机器人模型包括工业机械臂、移动机器人、人形机器人等多种类型的精确三维模型及其运动学参数。传感器数据提供激光雷达、深度相机、IMU等虚拟传感器的配置文件和标定参数。环境场景包含室内外仿真环境（如仓库、工厂、城市街道）的三维网格、纹理及光照设置。物理属性数据涉及碰撞体配置、材料摩擦系数、刚体动力学参数等物理引擎所需的元数据。所有资产均采用标准化格式，支持直接导入Isaac Sim进行机器人感知、导航、抓取等任务的仿真训练，并具备与真实世界数据对接的接口规范。

This dataset contains digital assets for robotic simulation and artificial intelligence training on the NVIDIA Isaac Sim platform, covering four core categories of data: robotic models, sensor resources, environment scenes, and physical property metadata. Robotic models include precise 3D models and their accompanying kinematic parameters for various robot types such as industrial robotic arms, mobile robots, and humanoid robots. Sensor resources provide configuration files and calibration parameters for virtual sensors including LiDAR, depth cameras, and IMUs. Environment scenes feature 3D meshes, textures, and lighting settings for both indoor and outdoor simulation environments like warehouses, factories, and urban streets. Physical property metadata includes data required by the physics engine, such as collider configurations, material friction coefficients, and rigid body dynamics parameters. All assets adopt standardized formats, supporting direct import into Isaac Sim for simulation training of robotic tasks including perception, navigation, and grasping, and are equipped with interface specifications for docking with real-world data.

搜集汇总

背景与挑战

背景概述

NVIDIA Isaac Sim Assets数据集包含机器人、传感器、材料和道具以及环境等资产，用于Isaac Sim仿真平台，支持机器人模拟和合成数据生成。