Update app.py

app.py

@@ -4,9 +4,16 @@ import torchaudio
 import cv2
 import os
 import numpy as np
+import tensorflow as tf
 
 emotion_labels = {0: 'neutral', 1: 'calm', 2: 'happy', 3: 'sad', 4: 'angry', 5: 'fearful'}
 
+def trained_model(model_path):
+    
+    model = load_model(model_path)
+
+    return model
+
 def process_video_audio(video_path, audio_path):
 
     wav = pt.tensor(list(audio_path[1]))
@@ -31,7 +38,6 @@ def process_video_audio(video_path, audio_path):
     
     cap = cv2.VideoCapture(video_path)
     frame_idx = 0
-    last_frame = None
     for i in range(100):
         ret, frame = cap.read()
         if ret and (i % 10 == 0):
@@ -45,16 +51,15 @@ def process_video_audio(video_path, audio_path):
             else:
                 resized_frame = cv2.resize(frame, (120, 120))
                 train_visual[0, :, :, :, frame_idx] = pt.tensor(resized_frame)
-            last_frame = frame
             frame_idx += 1
     cap.release()
-    predicted_emotion = "unknown"
 
-    return last_frame
+    return train_visual, train_audio_wave, train_audio_cnn
 
 # Định nghĩa giao diện Gradio
 def gradio_interface(video, audio):
-    frame= process_video_audio(video, audio)
+    train_visual, train_audio_wave, train_audio_cnn = process_video_audio(video, audio)
+    model = trained_model("./model_vui_ve.h5")
     return frame
 
 iface = gr.Interface(