Ev-Layout
收藏资源简介:
Ev-Layout是一个专为室内布局估计和跟踪设计的大型事件驱动多模态数据集。该数据集通过一个头戴式数据采集平台,整合了RGB相机和事件相机,能够在运动中捕获室内布局。数据集收集过程中还记录了IMU时间序列数据和环境亮度信息,以突出运动速度和光照对布局估计精度的影响。Ev-Layout数据集包含2500个序列,超过771,300张RGB图像和100亿个事件数据点,其中39,000张图像标注了室内布局,以支持事件驱动或基于视频的室内布局估计研究。
Ev-Layout is a large-scale event-driven multimodal dataset specifically designed for indoor layout estimation and tracking. It integrates RGB and event cameras via a head-mounted data acquisition platform, enabling the capture of indoor layouts during movement. During data collection, IMU time-series data and ambient brightness information are also recorded to highlight the impacts of movement speed and lighting conditions on the accuracy of layout estimation. The Ev-Layout dataset includes 2500 sequences, over 771,300 RGB images and 10 billion event data points, among which 39,000 images are annotated with indoor layouts to support research on event-driven or video-based indoor layout estimation.

- 1Ev-Layout: A Large-scale Event-based Multi-modal Dataset for Indoor Layout Estimation and Tracking山东大学软件学院, 新加坡南洋理工大学电子与电气工程学院 · 2025年



