Update app.py

app.py

@@ -3,9 +3,6 @@ import cv2
 import mediapipe as mp
 import tempfile
 
-from micro_gestures import thumb_up, thumb_down, index_pointing_up, fist_closed, palm_open, ok_sign
-from composite_gestures import detect_composite_gesture
-
 # initialize mediapipe modules
 mp_hands = mp.solutions.hands
 mp_pose = mp.solutions.pose
@@ -32,6 +29,8 @@ def process_video(video_path, target_width=640):
     temp_output = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4")
     out = None
 
+    last_label = "None"  # store last detected gesture label
+
     while cap.isOpened():
         ret, frame = cap.read()
         if not ret:
@@ -47,7 +46,6 @@ def process_video(video_path, target_width=640):
         rgb_small = cv2.cvtColor(frame_small, cv2.COLOR_BGR2RGB)
 
         # hand detection
-        detected_micro = []
         hand_results = hands.process(rgb_small)
         if hand_results.multi_hand_landmarks:
             for hand_landmarks in hand_results.multi_hand_landmarks:
@@ -55,22 +53,9 @@ def process_video(video_path, target_width=640):
                     frame_small,
                     hand_landmarks,
                     mp_hands.HAND_CONNECTIONS,
-                    mp_drawing.DrawingSpec(color=(0,0,255), thickness=1, circle_radius=1),  # hand landmarks
-                    mp_drawing.DrawingSpec(color=(0,255,0), thickness=1, circle_radius=1)   # hand connections
+                    mp_drawing.DrawingSpec(color=(0,0,255), thickness=1, circle_radius=1),
+                    mp_drawing.DrawingSpec(color=(0,255,0), thickness=1, circle_radius=1)
                 )
-                # check micro gestures
-                if thumb_up(hand_landmarks.landmark):
-                    detected_micro.append("thumb_up")
-                if thumb_down(hand_landmarks.landmark):
-                    detected_micro.append("thumb_down")
-                if index_pointing_up(hand_landmarks.landmark):
-                    detected_micro.append("index_pointing_up")
-                if fist_closed(hand_landmarks.landmark):
-                    detected_micro.append("fist_closed")
-                if palm_open(hand_landmarks.landmark):
-                    detected_micro.append("palm_open")
-                if ok_sign(hand_landmarks.landmark):
-                    detected_micro.append("ok_sign")
 
         # pose detection
         pose_results = pose.process(rgb_small)
@@ -79,22 +64,19 @@ def process_video(video_path, target_width=640):
                 frame_small,
                 pose_results.pose_landmarks,
                 mp_pose.POSE_CONNECTIONS,
-                mp_drawing.DrawingSpec(color=(255,0,0), thickness=1, circle_radius=1),  # body landmarks
-                mp_drawing.DrawingSpec(color=(0,255,255), thickness=1, circle_radius=1) # body connections
+                mp_drawing.DrawingSpec(color=(255,0,0), thickness=1, circle_radius=1),
+                mp_drawing.DrawingSpec(color=(0,255,255), thickness=1, circle_radius=1)
             )
 
-        # determine composite gesture
-        composite_label = detect_composite_gesture(detected_micro)
-        if composite_label:
-            # put text label on the frame
-            cv2.putText(frame_small, composite_label, (10, target_height-10),
-                        cv2.FONT_HERSHEY_SIMPLEX, 1, (0,0,255), 2, cv2.LINE_AA)
+        # here you would detect gesture label (example placeholder)
+        # last_label = detect_composite_gesture([...])
+        last_label = "example_label"
 
         # initialize video writer
         if out is None:
             fps = cap.get(cv2.CAP_PROP_FPS)
-            if fps <= 0 or fps > 120:  # if invalid or absurd
-                fps = 30  # default fallback
+            if fps <= 0 or fps > 120:
+                fps = 30
             
             out = cv2.VideoWriter(
                 temp_output.name,
@@ -110,7 +92,8 @@ def process_video(video_path, target_width=640):
     if out:
         out.release()
 
-    return temp_output.name
+    # return both video path and last label for gradio
+    return temp_output.name, last_label
 
 # gradio interface
 iface = gr.Interface(
@@ -119,9 +102,12 @@ iface = gr.Interface(
         gr.Video(label="Upload or Record Video"),
         gr.Slider(minimum=160, maximum=1280, value=640, step=20, label="Processing Width")
     ],
-    outputs=gr.Video(label="Processed Video with Landmarks"),
+    outputs=[
+        gr.Video(label="Processed Video with Landmarks"),
+        gr.Textbox(label="Detected Gesture", interactive=False)
+    ],
     title="Hand & Body Pose Detection",
-    description="Upload or record a video, and see MediaPipe detect hand and body landmarks with connections and detected gesture labels."
+    description="Upload or record a video, and see MediaPipe detect hand and body landmarks with connections."
 )
 
 if __name__ == "__main__":