Update README.md
Browse files
README.md
CHANGED
|
@@ -37,4 +37,69 @@ license: cc-by-nc-nd-4.0
|
|
| 37 |
---
|
| 38 |
|
| 39 |
license: cc-by-nc-nd-4.0
|
| 40 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
---
|
| 38 |
|
| 39 |
license: cc-by-nc-nd-4.0
|
| 40 |
+
---
|
| 41 |
+
|
| 42 |
+
# BATIS: Benchmarking Bayesian Approaches for Improving Species Distribution Models
|
| 43 |
+
|
| 44 |
+
This repository contains the dataset used in experiments shown in BATIS: Benchmarking Bayesian Approaches for Improving Species Distribution Models (preprint).
|
| 45 |
+
|
| 46 |
+
## Dataset Configurations and Splits
|
| 47 |
+
|
| 48 |
+
The dataset contains the following four configurations :
|
| 49 |
+
|
| 50 |
+
- **Kenya :** Containing the data used to train our models for predicting bird species distribution in Kenya.
|
| 51 |
+
- **South Africa :** Containing the data used to train our models for predicting bird species distribution in South Africa.
|
| 52 |
+
- **USA-Winter :** Containing the data used to train our models for predicting bird species distribution in the United States of America during the winter season.
|
| 53 |
+
**USA-Summer :** Containing the data used to train our models for predicting bird species distribution in the United States of America during the summer season.
|
| 54 |
+
|
| 55 |
+
Each subset can be further divided into `train`, `test` and `split`. These splits are the same as the one we used in our paper, and were generated by following the pre-processing pipeline described in our paper, which can be easily reproduced by re-using our [https://github.com/cath34/batis_benchmark_neurips_2025/blob/main/Splits/combined_script.py](code).
|
| 56 |
+
|
| 57 |
+
## Dataset Structure
|
| 58 |
+
|
| 59 |
+
```
|
| 60 |
+
/batis_benchmark_2025/
|
| 61 |
+
Kenya/
|
| 62 |
+
images.tar.gz
|
| 63 |
+
environmental.tar.gz
|
| 64 |
+
targets.tar.gz
|
| 65 |
+
species_list.csv
|
| 66 |
+
train_filtered.csv
|
| 67 |
+
test_filtered.csv
|
| 68 |
+
valid_filtered.csv
|
| 69 |
+
South Africa/
|
| 70 |
+
images.tar.gz
|
| 71 |
+
environmental.tar.gz
|
| 72 |
+
targets.tar.gz
|
| 73 |
+
species_list.csv
|
| 74 |
+
train_filtered.csv
|
| 75 |
+
test_filtered.csv
|
| 76 |
+
valid_filtered.csv
|
| 77 |
+
USA-Winter/
|
| 78 |
+
images/
|
| 79 |
+
images_{aa}
|
| 80 |
+
...
|
| 81 |
+
images_{ad}
|
| 82 |
+
|
| 83 |
+
environmental.tar.gz
|
| 84 |
+
targets.tar.gz
|
| 85 |
+
species_list.csv
|
| 86 |
+
train_filtered.csv
|
| 87 |
+
test_filtered.csv
|
| 88 |
+
valid_filtered.csv
|
| 89 |
+
USA-Summer/
|
| 90 |
+
images.tar.gz
|
| 91 |
+
environmental.tar.gz
|
| 92 |
+
targets.tar.gz
|
| 93 |
+
species_list.csv
|
| 94 |
+
train_filtered.csv
|
| 95 |
+
test_filtered.csv
|
| 96 |
+
valid_filtered.csv
|
| 97 |
+
|
| 98 |
+
```
|
| 99 |
+
|
| 100 |
+
## Data Fields
|
| 101 |
+
|
| 102 |
+
## Authors
|
| 103 |
+
|
| 104 |
+
## Acknowledgments
|
| 105 |
+
|