cs-chan/Exclusively-Dark-Image-Dataset
Exclusively Dark (ExDARK) dataset which to the best of our knowledge, is the largest collection of low-light images taken in very low-light environments to twilight (i.e 10 different conditions) to-date with image class and object level annotations.
What it solves
It provides a specialized dataset to help researchers improve object detection and image enhancement in low-light environments, where standard datasets often fail due to poor visibility.
How it works
The project provides the Exclusively Dark (ExDark) dataset, which consists of 7,363 images ranging from twilight to very low-light conditions. These images cover 10 different lighting conditions and 12 object classes, with annotations provided at both the image class level and through local object bounding boxes.
Who it’s for
Computer vision researchers and developers working on low-light image processing, object detection, and image enhancement algorithms.
Highlights
- 7,363 low-light images across 10 different lighting conditions.
- 12 object classes annotated with bounding boxes.
- Includes source code for low-light image enhancement (SPIC).
Related
- Project
- Project
- Project
- Project