Update README.md

README.md

@@ -1,6 +1,12 @@
-This folder contains samples of existing image classification/OOD robustness
-benchmarks that have been assigned labeles for a natural vs. rendition
-classification tasks (i.e., natural, ambigious, or rendition).
+---
+license: cc-by-4.0
+pretty_name: ImageNet-DomainNet-NvR
+---
 
-These folders here should be used only to test new rendition
-classifiers and not to train/design new ones!
+This dataset supplies new classification labels to existing image classification/OOD robustness benchmarks.
+The classification task is about distinguishing images with a natural style (e.g., photographs) from renditions (e.g., sketches, cartoons).
+There exist three classes: "natural", "ambiguous", and "rendition".
+
+These samples should be used only to test new rendition classifiers and not to train/design new ones!
+
+When using this dataset, please also cite the work introducing each of the used data sources (see README.md in each .tar file).