Update README.md

README.md

@@ -17,4 +17,12 @@ language:
 - ja
 size_categories:
 - 1K<n<10K
----
+---
+
+
+# Dataset Summary
+
+The expanded XTD-11 dataset, now including Arabic, enhances the original XTD collection. This dataset introduces a 1,000-image multi-lingual MSCOCO2014 caption to test multimodel in zeroshot image or text retrieval in 11 Languages.
+
+
+# Dataset