Datasets:

Modalities:
Image
Text
Formats:
parquet
Languages:
English
ArXiv:
License:
gaslen commited on
Commit
747d0c6
·
verified ·
1 Parent(s): 2a9f98b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -13
README.md CHANGED
@@ -73,7 +73,14 @@ Valerie Zermatten · Javiera Castillo-Navarro · Pallavi Jain · Devis Tuia · D
73
  [![GitHub](https://img.shields.io/badge/GitHub-Code-black)](https://github.com/eceo-epfl/EcoWikiRS)
74
  [![DOI](https://img.shields.io/badge/DOI-Zenodo-blue)](https://zenodo.org/records/15236742)
75
 
76
- ## Two data formats in this repository
 
 
 
 
 
 
 
77
 
78
  ### 1. Original raw data
79
  - In`ecowikirs.zip`
@@ -82,28 +89,20 @@ Valerie Zermatten · Javiera Castillo-Navarro · Pallavi Jain · Devis Tuia · D
82
 
83
  ### 2. Hugging Face processed version
84
  - In the `data/` directory
85
- - Used automatically by `load_dataset("EPFL-ECEO/EcoWikiRS")`
86
 
87
- ### Features
88
  - id (string): Unique site/sample identifier
89
  - EUNIS (ClassLabel): Habitat class label. Original codes remapped to contiguous indices to handle gaps.
90
- species_list (list of strings): All species recorded at the site.
91
  - image (Image): PNG version of original TIFF image (from `swissimage/{id}.tif` in the raw data), loaded as PIL image.
92
- - species_text (json): Aggregated JSON data per species (from `wiki/{species_name}.json` in the raw data), keyed by binomial_name, including keywords, habitat_section, and random_sentences.
93
 
94
  ### Notes
95
  EUNIS labels have been remapped in the Hugging Face processed version to contiguous integers. The mapping is the following:
96
  ```{0: 14, 1: 15, 2: 16, 3: 21, 4: 23, 5: 24, 6: 25, 7: 26, 8: 31, 9: 32, 10: 40, 11: 41, 12: 43, 13: 44, 14: 45, 15: 47, 16: 48, 17: 49, 18: 50, 19: 52, 20: 53, 21: 54, 22: 55, 23: 56, 24: 57}```
97
 
98
 
99
- ## Overview
100
-
101
- The **WikiRS** dataset, composed of triplets of images, species list and Wikipedia sentences :
102
- - **91k high-resolution aerial images** (50cm, RGB bands) from the swissIMAGE product
103
- - crowd-sourced species observations from **2745 different species**, collected from **GBIF**, geolocated within the footprint of the aerial image.
104
- - **103K different sentences** describing the habitat of the observed species, extracted from the corresponding Wikipedia article.
105
-
106
-
107
  ## How to cite this work:
108
 
109
  ```
 
73
  [![GitHub](https://img.shields.io/badge/GitHub-Code-black)](https://github.com/eceo-epfl/EcoWikiRS)
74
  [![DOI](https://img.shields.io/badge/DOI-Zenodo-blue)](https://zenodo.org/records/15236742)
75
 
76
+ ## Overview
77
+
78
+ The **WikiRS** dataset, composed of triplets of images, species list and Wikipedia sentences :
79
+ - **91k high-resolution aerial images** (50cm, RGB bands) from the swissIMAGE product
80
+ - crowd-sourced species observations from **2745 different species**, collected from **GBIF**, geolocated within the footprint of the aerial image.
81
+ - **103K different sentences** describing the habitat of the observed species, extracted from the corresponding Wikipedia article.
82
+
83
+ ## Data formats in this repository
84
 
85
  ### 1. Original raw data
86
  - In`ecowikirs.zip`
 
89
 
90
  ### 2. Hugging Face processed version
91
  - In the `data/` directory
92
+ - Automatically loaded via `load_dataset("EPFL-ECEO/EcoWikiRS")`
93
 
94
+ ### HF Features
95
  - id (string): Unique site/sample identifier
96
  - EUNIS (ClassLabel): Habitat class label. Original codes remapped to contiguous indices to handle gaps.
97
+ species_list (list[string]): All species recorded at the site.
98
  - image (Image): PNG version of original TIFF image (from `swissimage/{id}.tif` in the raw data), loaded as PIL image.
99
+ - species_text (json string): Aggregated JSON data per species (from `wiki/{species_name}.json` in the raw data), keyed by binomial_name, including keywords, habitat_section, and random_sentences.
100
 
101
  ### Notes
102
  EUNIS labels have been remapped in the Hugging Face processed version to contiguous integers. The mapping is the following:
103
  ```{0: 14, 1: 15, 2: 16, 3: 21, 4: 23, 5: 24, 6: 25, 7: 26, 8: 31, 9: 32, 10: 40, 11: 41, 12: 43, 13: 44, 14: 45, 15: 47, 16: 48, 17: 49, 18: 50, 19: 52, 20: 53, 21: 54, 22: 55, 23: 56, 24: 57}```
104
 
105
 
 
 
 
 
 
 
 
 
106
  ## How to cite this work:
107
 
108
  ```