SDE dataset
收藏资源简介:
SDE数据集是由香港科技大学(广州)创建的大规模真实世界事件图像数据集,包含超过31,477对图像和事件,涵盖室内外低光和正常光照条件。数据集通过机器人手臂精确捕捉,确保空间对齐精度高达0.03mm,并采用匹配对齐策略,使90%的数据集时间对齐误差控制在0.01s以内。该数据集主要用于解决真实世界低光场景下的图像增强问题,特别是在需要高动态范围和高时间分辨率的场景中。
The SDE Dataset is a large-scale real-world event-image dataset developed by The Hong Kong University of Science and Technology (Guangzhou). It comprises over 31,477 paired image and event samples, covering both indoor and outdoor environments under both low-light and standard illumination conditions. It was precisely acquired using a robotic arm, guaranteeing a spatial alignment accuracy of up to 0.03 mm, and employs a matching alignment strategy to limit the temporal alignment error of 90% of the dataset to within 0.01 seconds. This dataset is primarily designed to tackle the challenge of image enhancement in real-world low-light scenarios, particularly those demanding high dynamic range and high temporal resolution.

- 1Towards Robust Event-guided Low-Light Image Enhancement: A Large-Scale Real-World Event-Image Dataset and Novel Approach香港科技大学(广州) · 2024年



