Add link to paper

README.md

@@ -1,53 +1,54 @@
 ---
+language:
+- en
+license: mit
+size_categories:
+- 100K<n<1M
+task_categories:
+- image-classification
 pretty_name: FedRS
 tags:
 - image
 - federated-learning
 - niid
 - classification
-license: mit
-language:
-- en
-task_categories:
-- image-classification
-size_categories:
-- 100K<n<1M
 dataset_size: 104914
 homepage: https://huggingface.co/datasets/LiberiOnTerra/FedRS
-
 configs:
-
-  - config_name: FedRS
-    data_files:
-      - split: train_NIID_1
-        path: FedRS/NIID-1/**/*.jpg
-      - split: train_NIID2
-        path: FedRS/NIID-2/**/*.jpg
-      - split: val_balanced
-        path: FedRS/val_balanced/**/*.jpg
-      - split: val_imbalanced
-        path: FedRS/val_imbalanced/**/*.jpg
-    features:
-      - name: image
-        dtype: image
-      - name: label
-        dtype: string
-
-  - config_name: FedRS_5
-    data_files:
-      - split: train_NIID_1
-        path: FedRS-5/NIID-1/**/*.jpg
-      - split: train_NIID2
-        path: FedRS-5/NIID-2/**/*.jpg
-      - split: val_balanced
-        path: FedRS-5/val_balanced/**/*.jpg
-      - split: val_imbalanced
-        path: FedRS-5/val_imbalanced/**/*.jpg
-    features:
-      - name: image
-        dtype: image
-      - name: label
-        dtype: string
+- config_name: FedRS
+  data_files:
+  - split: train_NIID_1
+    path: FedRS/NIID-1/**/*.jpg
+  - split: train_NIID2
+    path: FedRS/NIID-2/**/*.jpg
+  - split: val_balanced
+    path: FedRS/val_balanced/**/*.jpg
+  - split: val_imbalanced
+    path: FedRS/val_imbalanced/**/*.jpg
+  features:
+  - name: image
+    dtype: image
+  - name: label
+    dtype: string
+- config_name: FedRS_5
+  data_files:
+  - split: train_NIID_1
+    path: FedRS-5/NIID-1/**/*.jpg
+  - split: train_NIID2
+    path: FedRS-5/NIID-2/**/*.jpg
+  - split: val_balanced
+    path: FedRS-5/val_balanced/**/*.jpg
+  - split: val_imbalanced
+    path: FedRS-5/val_imbalanced/**/*.jpg
+  features:
+  - name: image
+    dtype: image
+  - name: label
+    dtype: string
 ---
+
 ## FedRS
+
+This dataset is based on the paper [FedRS-Bench: Realistic Federated Learning Datasets and Benchmarks in Remote Sensing](https://huggingface.co/papers/2505.08325).
+
 we obtain 135 clients in total. Among these clients, EuroSAT contributes 45 clients, AID 15 clients, RSD46-WHU 35 clients, NaSC-TG2 34 clients, NWPU-RESISC45 12 clients, AID 5 clients, UCM 2 clients and the remaining 2 clients are from WHU-RS19 and OPTIMAL-31. Each client has a non-overlapping subset of the data. The number of images per client ranges widely, from as few as 114 to 1559, mirroring the unbalanced nature of real federated data.