Upload dataset_card.yaml with huggingface_hub
Browse files- dataset_card.yaml +36 -0
dataset_card.yaml
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
pretty_name: UNICE Dataset
|
| 2 |
+
tags:
|
| 3 |
+
- image-enhancement
|
| 4 |
+
- multi-exposure
|
| 5 |
+
- low-light
|
| 6 |
+
- exposure-correction
|
| 7 |
+
- computer-vision
|
| 8 |
+
- pseudo-ground-truth
|
| 9 |
+
- contrast-enhancement
|
| 10 |
+
license: cc-by-nc-4.0
|
| 11 |
+
task_categories:
|
| 12 |
+
- image-enhancement
|
| 13 |
+
language:
|
| 14 |
+
- en
|
| 15 |
+
dataset_info:
|
| 16 |
+
features:
|
| 17 |
+
- name: input_image
|
| 18 |
+
dtype: image
|
| 19 |
+
- name: pseudo_gt
|
| 20 |
+
dtype: image
|
| 21 |
+
splits:
|
| 22 |
+
- name: train
|
| 23 |
+
num_bytes: ~
|
| 24 |
+
num_examples: ~
|
| 25 |
+
size_in_bytes: ~
|
| 26 |
+
dataset_type: image
|
| 27 |
+
annotations_creators:
|
| 28 |
+
- machine-generated
|
| 29 |
+
source_datasets:
|
| 30 |
+
- original
|
| 31 |
+
pretty_name: UNICE Multi-Exposure Dataset
|
| 32 |
+
description: >
|
| 33 |
+
The UNICE dataset is a large-scale dataset for universal image contrast enhancement.
|
| 34 |
+
It contains multi-exposure sequences (MES) rendered from HDR raw images and their corresponding
|
| 35 |
+
pseudo ground truths generated via multi-exposure fusion (MEF). The dataset supports training
|
| 36 |
+
models for low-light enhancement, exposure correction, backlit enhancement, and LDR-to-HDR transformation.
|