室内家装环境机器人训练场景数据

本数据集旨在为家居服务机器人提供全景式、精细化的感知与交互训练基础，支撑机器人在复杂、多样化的完整住宅环境中实现可靠导航、精准操作与高层任务规划。数据集构建了一个完整且真实的数字化家庭环境，覆盖客厅、餐厅、卧室、厨房、卫生间五大核心家居场景。场景内所有物体均经过实例级分割与清洗，确保模型边界清晰、无粘连；并进行了像素级语义标注，类别定义精细。这是常见家装场景，包含餐客厅，卧室，卫生间，厨房等，场景中实现了实例分割清洗，语义精准标注，部分模型提供了部件相关的信息。该场景我们使用了矩形灯光模拟真实的灯光，建议使用版本大于5的unreal仿真软件或版本为2022.1的issac sim仿真软件。

This dataset aims to provide a comprehensive and fine-grained perception and interaction training foundation for home service robots, supporting the robots to achieve reliable navigation, precise manipulation and high-level task planning in complex and diverse full residential environments. The dataset constructs a complete and realistic digitized home environment, covering five core home scenarios: living room, dining room, bedroom, kitchen and bathroom. All objects in the scenarios have undergone instance-level segmentation and cleaning to ensure clear and non-overlapping model boundaries, and have been annotated with pixel-level semantic labels with fine-grained category definitions. These are common home decoration scenarios, including dining-living room, bedroom, bathroom, kitchen and so on. The scenarios have completed instance segmentation and cleaning, precise semantic annotation, and some of the object models provide component-related information. Rectangular lighting is used in these scenarios to simulate real-world lighting. It is recommended to use Unreal simulation software with a version higher than 5 or Isaac Sim simulation software with version 2022.1.

搜集汇总

背景与挑战

背景概述

该数据集是一系列针对结直肠癌及胃部病变筛查的医疗检测数据，涵盖多个特定肠段（如横结肠、回肠、降结肠等）和胃部区域（如胃窦、胃体）的肠镜与胃镜检测结果。其核心特点是采用标准化的量化评分体系，基于病变部位、病理诊断、形态特征等多项指标，通过AHP层次分析法进行综合评分，将病变风险分为高、中、低三个等级，旨在为临床医生提供客观的风险评估、决策支持和结构化报告生成工具，以提升诊疗管理的标准化水平。