Refactor annotation handling and inference processing: Update 'time' attribute in Annotation class to private, streamline video time formatting in main processing function, and adjust annotations dictionary access in detection conversion. Ensure consistency in handling input frames during inference.

annotation.pxd

@@ -8,7 +8,7 @@ cdef class Detection:
 cdef class Annotation:
     cdef public str name
     cdef public str original_media_name
-    cdef public long time
+    cdef long time
     cdef public list[Detection] detections
     cdef public bytes image
 

constants_inf.pxd

@@ -21,7 +21,7 @@ cdef log(str log_message)
 cdef logerror(str error)
 cdef format_time(int ms)
 
-cdef dict annotations_dict
+cdef dict[int, AnnotationClass] annotations_dict
 
 cdef class AnnotationClass:
     cdef public int id

constants_inf.pyx

@@ -70,9 +70,6 @@ logger.add(
     colorize=True
 )
 
-def get_annotations_dict():
-    return annotations_dict
-
 cdef log(str log_message):
     logger.info(log_message)
 

inference.pyx

@@ -264,9 +264,7 @@ cdef class Inference:
         if frame is None:
             raise ValueError("Invalid image data")
 
-        cdef int bs = self.engine.get_batch_size()
-        frames = [frame] * bs
-        input_blob = self.preprocess(frames)
+        input_blob = self.preprocess([frame])
         outputs = self.engine.run(input_blob)
         list_detections = self.postprocess(outputs, ai_config)
         if list_detections:

main.py

@@ -109,10 +109,9 @@ class AIConfigDto(BaseModel):
 
 def detection_to_dto(det) -> DetectionDto:
     import constants_inf
-    ann = constants_inf.get_annotations_dict()
     label = ""
-    if det.cls in ann:
-        label = ann[det.cls].name
+    if det.cls in constants_inf.annotations_dict:
+        label = constants_inf.annotations_dict[det.cls].name
     return DetectionDto(
         centerX=det.x,
         centerY=det.y,
@@ -178,14 +177,10 @@ def _post_annotation_to_service(token_mgr: TokenManager, media_id: str,
     try:
         token = token_mgr.get_valid_token()
         image_b64 = base64.b64encode(annotation.image).decode() if annotation.image else None
-        total_seconds = annotation.time // 1000 if annotation.time else 0
-        hours = total_seconds // 3600
-        minutes = (total_seconds % 3600) // 60
-        seconds = total_seconds % 60
         payload = {
             "mediaId": media_id,
             "source": 0,
-            "videoTime": f"{hours:02d}:{minutes:02d}:{seconds:02d}",
+            "videoTime": f"00:00:{annotation.time // 1000:02d}" if annotation.time else "00:00:00",
             "detections": [d.model_dump() for d in dtos],
         }
         if image_b64: