Merge pull request #1092 from misrasaurabh1/codeflash/optimize-select_top_confidence_detection-m7drsd4o

⚡️ Speed up function `select_top_confidence_detection` by 188%

Author: PawelPeczek-Roboflow

inference/core/workflows/core_steps/common/query_language/operations/detections/base.py

@@ -106,11 +106,9 @@ def shift_detections(value: Any, shift_x: int, shift_y: int, **kwargs) -> sv.Det
 
 def select_top_confidence_detection(detections: sv.Detections) -> sv.Detections:
     if len(detections) == 0:
-        return deepcopy(detections)
-    confidence = detections.confidence
-    max_value = confidence.max()
-    index = np.argwhere(confidence == max_value)[0].item()
-    return detections[index]
+        return detections  # No need to deepcopy an empty detections object
+    max_index = np.argmax(detections.confidence)
+    return detections[max_index]
 
 
 def select_leftmost_detection(detections: sv.Detections) -> sv.Detections: