Tidy

app.py

@@ -35,21 +35,20 @@ def fn(video: gr.Video, actions: list[int]):
     actions = []
     detections = ([], [])
     for i, frame in enumerate(clip.iter_frames()):
-      if i % config.classify_action_frame_step == 0:
-        if do_classify:
+      if do_classify:
+        if i % config.classify_action_frame_step == 0:
           frames.append(format_frame(frame, config))
-      if i % config.detect_object_frame_step == 0:
-        print(f'Detect object: Frame {i}')
-        if do_detect:
-          detections = detect_object(detector, frame)
-      if len(frames) == config.classify_action_num_frames:
-        print(f'Classify action: Until frame {i}')
-        if do_classify:
+        if len(frames) == config.classify_action_num_frames:
+          print(f'Classify action: Until frame {i}')
           actions = classify_action(classifier, frames, config.id_to_name)
           frames = []
+      if do_detect:
+        if i % config.detect_object_frame_step == 0:
+          print(f'Detect object: Frame {i}')
+          detections = detect_object(detector, frame)
       if do_detect:
         frame = draw_boxes(frame, detections, actions, do_classify)
-      else:
+      elif len(actions) > 0:
         frame = draw_classes(frame, actions)
       processed_frames.append(frame)
       if i % config.yield_frame_steps == 0:

configuration.py

@@ -29,4 +29,4 @@ class Config:
   detect_object_frame_step = 5
   classify_action_frame_step = frame_step
   classify_action_num_frames = num_frames
-  yield_frame_steps = 3
+  yield_frame_steps = 2

inference.py

@@ -67,7 +67,7 @@ def draw_boxes(frame, detections, actions, do_classify):
 def draw_classes(frame, actions):
   height, width, _ = frame.shape
   labels = np.zeros((height, width), dtype=int)
-  label_names = ['Airplane\n' + '\n'.join(actions)]
+  label_names = ['\n'.join(actions)]
   frame = label2rgb(
     label=labels,
     image=frame,