Update README.md

RW-LVT is a large-scale real-world video dataset capturing low-visibility transportation scenarios including night, dusk, rain, small-town and city driving, and rail environments. The dataset emphasizes temporal structure and challenging perception conditions and is designed for self-supervised learning, robustness research, and world-model training.

README.md

@@ -1,3 +1,15 @@
----
-license: cc-by-4.0
----
+---
+license: cc-by-4.0
+task_categories:
+- video-classification
+- robotics
+- reinforcement-learning
+language:
+- en
+tags:
+- >-
+  video  driving  night  low-light  rain  transportation  robustness 
+  world-models  self-supervised  real-world
+size_categories:
+- 1K<n<10K
+---