semi-working

DeepEncode.py

@@ -16,7 +16,7 @@ from video_compression_model import VideoCompressionModel
 COMPRESSED_VIDEO_FILE = 'compressed_video.avi'
 MAX_FRAMES = 0  # Limit the number of frames processed
 CRF = 51
-SPEED = PRESET_SPEED_CATEGORIES.index("ultrafast")
+SPEED = PRESET_SPEED_CATEGORIES.index("veryslow")
 
 # Load the trained model
 MODEL = tf.keras.models.load_model('models/model.tf', custom_objects={'VideoCompressionModel': VideoCompressionModel})
@@ -30,21 +30,21 @@ def load_frame_from_video(video_file, frame_num):
     ret, frame = cap.read()
     if not ret:
         return None
-    #frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
     cap.release()
     
     return frame
 
 def predict_frame(uncompressed_frame):
 
-    #display_frame = np.clip(cv2.cvtColor(uncompressed_frame, cv2.COLOR_BGR2RGB) * 255.0, 0, 255).astype(np.uint8)
+    display_frame = np.clip(cv2.cvtColor(uncompressed_frame, cv2.COLOR_BGR2RGB) * 255.0, 0, 255).astype(np.uint8)
-    #cv2.imshow("uncomp", uncompressed_frame)
+    cv2.imshow("uncomp", uncompressed_frame)
     
     frame = preprocess_frame(uncompressed_frame, CRF, SPEED)
     
     compressed_frame = MODEL.predict([np.expand_dims(frame, axis=0)])[0]
+    compressed_frame = compressed_frame[:, :, :3]  # Keep only the first 3 channels (BGR)
     
-    display_frame = np.clip(cv2.cvtColor(compressed_frame, cv2.COLOR_BGR2RGB) * 255.0, 0, 255).astype(np.uint8)
+    display_frame = np.clip(compressed_frame * 255.0, 0, 255).astype(np.uint8)
     
     cv2.imshow("comp", display_frame)
     cv2.waitKey(1)

test_data/validation/validation.json

@@ -22,5 +22,29 @@
         "original_video_file": "Scene5.mkv",
         "crf": 51,
         "preset_speed": "veryslow"
+    },
+    {
+        "compressed_video_file": "Scene6_x264_crf-25_preset-medium.mkv",
+        "original_video_file": "Scene6.mkv",
+        "crf": 25,
+        "preset_speed": "medium"
+    },
+    {
+        "compressed_video_file": "Scene7_x264_crf-34_preset-slower.mkv",
+        "original_video_file": "Scene7.mkv",
+        "crf": 34,
+        "preset_speed": "slower"
+    },
+    {
+        "compressed_video_file": "Scene8_x264_crf-12_preset-faster.mkv",
+        "original_video_file": "Scene8.mkv",
+        "crf": 12,
+        "preset_speed": "faster"
+    },
+    {
+        "compressed_video_file": "Scene9_x264_crf-15_preset-slow.mkv",
+        "original_video_file": "Scene9.mkv",
+        "crf": 15,
+        "preset_speed": "slow"
     }
 ]

train_model.py

@@ -16,7 +16,7 @@ from globalVars import HEIGHT, WIDTH, MAX_FRAMES, LOGGER
 # Constants
 BATCH_SIZE = 16
 EPOCHS = 100
-LEARNING_RATE = 0.001
+LEARNING_RATE = 0.000001
 MODEL_SAVE_FILE = "models/model.tf"
 MODEL_CHECKPOINT_DIR = "checkpoints"
 EARLY_STOP = 10

video_compression_model.py

@@ -19,8 +19,9 @@ def data_generator(videos, batch_size):
         # Iterate over each video
         for video_details in videos:
             # Get the paths for compressed and original (uncompressed) video files
-            video_path = os.path.join(os.path.dirname("test_data/validation/validation.json"), video_details["compressed_video_file"])
+            base_dir = os.path.dirname("test_data/validation/validation.json")
-            uncompressed_video_path = os.path.join(os.path.dirname("test_data/validation/validation.json"), video_details["original_video_file"])
+            video_path = os.path.join(base_dir, video_details["compressed_video_file"])
+            uncompressed_video_path = os.path.join(base_dir, video_details["original_video_file"])
             
             CRF = video_details["crf"] / 51
             SPEED = PRESET_SPEED_CATEGORIES.index(video_details["preset_speed"])
@@ -87,12 +88,14 @@ class VideoCompressionModel(tf.keras.Model):
             tf.keras.layers.UpSampling2D((2, 2)),
             tf.keras.layers.Conv2DTranspose(64, (3, 3), activation='relu', padding='same'),
             tf.keras.layers.UpSampling2D((2, 2)),
-            tf.keras.layers.Conv2DTranspose(1, (3, 3), activation='sigmoid', padding='same')
+            tf.keras.layers.Conv2DTranspose(NUM_COLOUR_CHANNELS + 2, (3, 3), activation='sigmoid', padding='same')
         ])
 
     def call(self, inputs):
-        # Encode the input
+        #print("Input shape:", inputs.shape)
         encoded = self.encoder(inputs)
-        # Decode the encoded representation
+        #print("Encoded shape:", encoded.shape)
         decoded = self.decoder(encoded)
+        #print("Decoded shape:", decoded.shape)
         return decoded
+