Update app.py

app.py

@@ -4,7 +4,7 @@ import numpy as np
 import tensorflow as tf
 import os
 
-# Load the trained models using Keras
+# Load the trained models
 cnn_model = tf.keras.models.load_model('noq_c_model.h5')
 qcnn_model = tf.keras.models.load_model('q_model.h5')
 
@@ -22,16 +22,19 @@ def get_video_paths(directory):
             videos.append(os.path.join(directory, vid))
     return videos
 
-# Get video paths for each category
-original_videos = get_video_paths(original_dir)
-deepfake_roop_videos = get_video_paths(deepfake_roop_dir)
-deepfake_web_videos = get_video_paths(deepfake_web_dir)
-
-# Combine all examples
-examples = original_videos + deepfake_roop_videos + deepfake_web_videos
+# Get example videos for each category
+examples_original = get_video_paths(original_dir)
+examples_deepfake_roop = get_video_paths(deepfake_roop_dir)
+examples_deepfake_web = get_video_paths(deepfake_web_dir)
 
 # Map from example video path to label
-example_videos_dict = {example['path']: example['label'] for example in examples}
+example_videos_dict = {}
+for vid in examples_original:
+    example_videos_dict[vid] = 'Original'
+for vid in examples_deepfake_roop:
+    example_videos_dict[vid] = 'DeepfakeRoop'
+for vid in examples_deepfake_web:
+    example_videos_dict[vid] = 'DeepfakeWeb'
 
 def process_video(video_path, true_label=None):
     cap = cv2.VideoCapture(video_path)
@@ -41,7 +44,7 @@ def process_video(video_path, true_label=None):
     frame_interval = max(int(round(fps / 30)), 1)
     frame_count = 0
     sampled_frames = []
-    
+
     while cap.isOpened():
         ret, frame = cap.read()
         if not ret:
@@ -54,7 +57,7 @@ def process_video(video_path, true_label=None):
             sampled_frames.append(normalized_frame)
         frame_count += 1
     cap.release()
-    
+
     cnn_correct = 0
     qcnn_correct = 0
     cnn_class0 = 0
@@ -71,7 +74,7 @@ def process_video(video_path, true_label=None):
             cnn_class1 += 1
         if true_label is not None and cnn_label == true_label:
             cnn_correct += 1
-        frame = frame.reshape(1, -1)  # Flatten the frame
+
         qcnn_pred = qcnn_model.predict(frame)
         qcnn_label = np.argmax(qcnn_pred)
         if qcnn_label == 0:
@@ -80,7 +83,7 @@ def process_video(video_path, true_label=None):
             qcnn_class1 += 1
         if true_label is not None and qcnn_label == true_label:
             qcnn_correct += 1
-    
+
     if total_frames > 0:
         cnn_class0_percent = (cnn_class0 / total_frames) * 100
         cnn_class1_percent = (cnn_class1 / total_frames) * 100
@@ -88,7 +91,7 @@ def process_video(video_path, true_label=None):
         qcnn_class1_percent = (qcnn_class1 / total_frames) * 100
     else:
         cnn_class0_percent = cnn_class1_percent = qcnn_class0_percent = qcnn_class1_percent = 0
-    
+
     if true_label is not None:
         # Calculate accuracy if true_label is provided (example video)
         cnn_accuracy = (cnn_correct / total_frames) * 100 if total_frames > 0 else 0
@@ -110,7 +113,7 @@ def predict(video_input):
         video_path = video_input
     else:
         return "Invalid video input."
-    
+
     # Check if video is an example
     if video_path in example_videos_dict:
         label = example_videos_dict[video_path]
@@ -125,30 +128,32 @@ def predict(video_input):
     return result
 
 with gr.Blocks() as demo:
-    gr.HTML("<h1 style='text-align: center;'>Quanvolutional Neural Networks for Deepfake Detection</h1>")
-    gr.HTML("<h2 style='text-align: center;'>Steven Fernandes, Ph.D.</h2>")
-    
-    with gr.Row():
+    gr.Markdown("<h1 style='text-align: center;'>Quanvolutional Neural Networks for Deepfake Detection</h1>")
+    gr.Markdown("<h2 style='text-align: center;'>Steven Fernandes, Ph.D.</h2>")
+
+    with gr.Row(elem_classes="content-container"):
         with gr.Column():
-            video_input = gr.Video(label="Upload Video",)
+            video_input = gr.Video(label="Upload Video", type="filepath")
+            with gr.Row(elem_classes="control-panel"):
+                pass  # Add any controls if needed
             examples_original = gr.Examples(
                 label="Original Videos",
                 inputs=video_input,
-                examples=original_videos
+                examples=examples_original,
             )
             examples_deepfake_roop = gr.Examples(
                 label="Deepfake Roop Videos",
                 inputs=video_input,
-                examples=deepfake_roop_videos
+                examples=examples_deepfake_roop,
             )
             examples_deepfake_web = gr.Examples(
                 label="Deepfake Web Videos",
                 inputs=video_input,
-                examples=deepfake_web_videos
+                examples=examples_deepfake_web,
             )
         with gr.Column():
             output = gr.Textbox(label="Result")
-            predict_button = gr.Button("Predict")
-    
+            predict_button = gr.Button("Predict", elem_classes="gr-button")
+
     predict_button.click(fn=predict, inputs=video_input, outputs=output)
     demo.launch()