jfloresf commited on
Commit
85e8cbd
1 Parent(s): e835cb3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +91 -3
README.md CHANGED
@@ -4,11 +4,99 @@ language:
4
  tags:
5
  - clouds
6
  - sentinel-2
 
 
7
  - remote-sensing
8
- pretty_name: cloudSEN12
9
  ---
10
- # cloudSEN12
11
 
12
  ***``A dataset about clouds from Sentinel-2``***
13
 
14
- cloudSEN12 is a dataset about clouds from Sentinel-2. It contains 4 classes: clear, thin, thick and overcast. The dataset is composed of 1800 images of 64x64 pixels. The dataset is divided into 3 splits: train, validation and test.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
  tags:
5
  - clouds
6
  - sentinel-2
7
+ - image-segmentation
8
+ - deep-learning
9
  - remote-sensing
10
+ pretty_name: cloudsen12
11
  ---
12
+ # cloudsen12
13
 
14
  ***``A dataset about clouds from Sentinel-2``***
15
 
16
+ CloudSEN12 is a LARGE dataset (~1 TB) for cloud semantic understanding that consists of 49,400 image patches (IP) that are evenly spread throughout all continents except Antarctica. Each IP covers 5090 x 5090 meters and contains data from Sentinel-2 levels 1C and 2A, hand-crafted annotations of thick and thin clouds and cloud shadows, Sentinel-1 Synthetic Aperture Radar (SAR), digital elevation model, surface water occurrence, land cover classes, and cloud mask results from six cutting-edge cloud detection algorithms.
17
+ CloudSEN12 is designed to support both weakly and self-/semi-supervised learning strategies by including three distinct forms of hand-crafted labeling data: high-quality, scribble and no-annotation. For more details on how we created the dataset see our paper: CloudSEN12 - a global dataset for semantic understanding of cloud and cloud shadow in Sentinel-2.
18
+
19
+
20
+ **ML-STAC Snippet**
21
+ ```python
22
+ import mlstac
23
+ secret = 'https://huggingface.co/datasets/jfloresf/mlstac-demo/resolve/main/main.json'
24
+ train_db = mlstac.load(secret, framework='torch', stream=True, device='cpu')
25
+ ```
26
+
27
+ **Sensor: Sentinel 2 - MSI**
28
+
29
+ **ML-STAC Task: TensorToTensor, TensorSegmentation**
30
+
31
+ **Data raw repository: [http://www.example.com/](http://www.example.com/)**
32
+
33
+ **Dataset discussion: [https://github.com/IPL-UV/ML-STAC/discussions/2](https://github.com/IPL-UV/ML-STAC/discussions/2)**
34
+
35
+ **Review mean score: 5.0**
36
+
37
+ **Split_strategy: random**
38
+
39
+ **Paper: [https://www.nature.com/articles/s41597-022-01878-2](https://www.nature.com/articles/s41597-022-01878-2)**
40
+ ## Data Providers
41
+
42
+ |Name|Role|URL|
43
+ | :---: | :---: | :---: |
44
+ |Image & Signal Processing|['host']|https://isp.uv.es/|
45
+ |ESA|['producer']|https://www.esa.int/|
46
+
47
+ ## Curators
48
+
49
+ |Name|Organization|URL|
50
+ | :---: | :---: | :---: |
51
+ |Cesar Aybar|Image & Signal Processing|http://csaybar.github.io/|
52
+
53
+ ## Reviewers
54
+
55
+ |Name|Organization|URL|Score|
56
+ | :---: | :---: | :---: | :---: |
57
+ |Cesar Aybar|Image & Signal Processing|http://csaybar.github.io/|5|
58
+
59
+ ## Labels
60
+
61
+ |Name|Value|
62
+ | :---: | :---: |
63
+ |clear|0|
64
+ |thick-cloud|1|
65
+ |thin-cloud|2|
66
+ |cloud-shadow|3|
67
+
68
+ ## Dimensions
69
+
70
+ ### input
71
+
72
+ |Axis|Name|Description|
73
+ | :---: | :---: | :---: |
74
+ |0|C|Channels - Spectral bands|
75
+ |1|H|Height|
76
+ |2|W|Width|
77
+
78
+ ### target
79
+
80
+ |Axis|Name|Description|
81
+ | :---: | :---: | :---: |
82
+ |0|C|Hand-crafted labels|
83
+ |1|H|Height|
84
+ |2|W|Width|
85
+
86
+ ## Spectral Bands
87
+
88
+ |Name|Common Name|Description|Center Wavelength|Full Width Half Max|Index|
89
+ | :---: | :---: | :---: | :---: | :---: | :---: |
90
+ |B01|coastal aerosol|Band 1 - Coastal aerosol - 60m|443.5|17.0|0|
91
+ |B02|blue|Band 2 - Blue - 10m|496.5|53.0|1|
92
+ |B03|green|Band 3 - Green - 10m|560.0|34.0|2|
93
+ |B04|red|Band 4 - Red - 10m|664.5|29.0|3|
94
+ |B05|red edge 1|Band 5 - Vegetation red edge 1 - 20m|704.5|13.0|4|
95
+ |B06|red edge 2|Band 6 - Vegetation red edge 2 - 20m|740.5|13.0|5|
96
+ |B07|red edge 3|Band 7 - Vegetation red edge 3 - 20m|783.0|18.0|6|
97
+ |B08|NIR|Band 8 - Near infrared - 10m|840.0|114.0|7|
98
+ |B8A|red edge 4|Band 8A - Vegetation red edge 4 - 20m|864.5|19.0|8|
99
+ |B09|water vapor|Band 9 - Water vapor - 60m|945.0|18.0|9|
100
+ |B10|cirrus|Band 10 - Cirrus - 60m|1375.5|31.0|10|
101
+ |B11|SWIR 1|Band 11 - Shortwave infrared 1 - 20m|1613.5|89.0|11|
102
+ |B12|SWIR 2|Band 12 - Shortwave infrared 2 - 20m|2199.5|173.0|12|