Add link to paper and code

README.md

@@ -1,3 +1,13 @@
+---
+license: mit
+tags:
+- video-understanding
+---
+
+Dataset of the paper [Interacted Object Grounding in Spatio-Temporal Human-Object Interactions](https://huggingface.co/papers/2412.19542)
+
+Code: https://github.com/DirtyHarryLYL/HAKE-AVA
+
 # Preparing Dataset for HAKE-GIO + HAKE-AVA-PaSta (h box + b box & class + action + PaSta)
 
 1. Dataset downloading steps
@@ -57,9 +67,3 @@
     - object_name: name of object
     - human_id: ID of the person performing the action
     - object_id: category id of the object
-
-
-
----
-license: mit
----