SatwikKambham/ex-dark
收藏资源简介:
Exclusively Dark (ExDARK) 数据集是一个包含在极低光环境下拍摄的图像集合,涵盖了从极低光到黄昏的10种不同光照条件。数据集中的图像标注了12个对象类别,类似于PASCAL VOC数据集,标注信息包括图像类别级别和局部对象的边界框。对象类别包括:狗、摩托车、人、猫、椅子、桌子、汽车、自行车、瓶子、公交车、杯子和船。数据集的特征包括图像、标签和边界框。数据集分为训练集,包含7361个样本,总大小为1770065字节。
Exclusively Dark (ExDARK) dataset is a collection of images captured under extremely low-light environments, covering 10 distinct illumination conditions ranging from extreme low light to dusk. The images in this dataset are annotated with 12 object categories, similar to the PASCAL VOC dataset, and the annotation information includes image-level category labels and bounding boxes for local objects. The 12 object categories are: dog, motorcycle, person, cat, chair, table, car, bicycle, bottle, bus, cup, and boat. The dataset consists of images, labels, and bounding boxes. The dataset is split into a training set containing 7361 samples with a total size of 1770065 bytes.
Exclusively Dark (ExDARK) 数据集
概述
Exclusively Dark (ExDARK) 数据集是一个包含低光照环境下图像的数据集,涵盖从极低光照到黄昏(共10种不同条件),并标注了12个对象类别的图像类别和局部对象边界框。
特征
- 图像 (img): 数据类型为图像。
- 标签 (labels): 包含以下类别:
- Dog
- Motorbike
- People
- Cat
- Chair
- Table
- Car
- Bicycle
- Bottle
- Bus
- Cup
- Boat
- 边界框 (bboxes): 数据类型为 float32,长度为4。
数据分割
- 训练集 (train): 包含7361个样本,大小为1770065字节。
数据集大小
- 下载大小: 1487935964字节
- 数据集大小: 1770065字节




