[fix] bug - criticial for pandaset use case

Author: chyomin06

compressai_vision/model_wrappers/detectron2.py

@@ -611,7 +611,11 @@ def __init__(self, device="cpu", **kwargs):
 
     @torch.no_grad()
     def _feature_pyramid_to_output(
-        self, x: Dict, org_img_size: Dict, input_img_size: List
+        self,
+        x: Dict,
+        org_img_size: Dict,
+        input_img_size: List,
+        hyper_params: Optional[Dict] = None,
     ):
         """
         performs  downstream task using the feature pyramid ['p2', 'p3', 'p4', 'p5']