test
This commit is contained in:
parent
8c5001166d
commit
5085c87300
3 changed files with 96 additions and 173 deletions
131
train_model.py
131
train_model.py
|
|
@ -1,16 +1,16 @@
|
|||
import os
|
||||
import json
|
||||
import tensorflow as tf
|
||||
import numpy as np
|
||||
import cv2
|
||||
from video_compression_model import NUM_CHANNELS, NUM_FRAMES, VideoCompressionModel, PRESET_SPEED_CATEGORIES
|
||||
import tensorflow as tf
|
||||
from video_compression_model import NUM_CHANNELS, VideoCompressionModel, PRESET_SPEED_CATEGORIES
|
||||
from tensorflow.keras.callbacks import EarlyStopping
|
||||
|
||||
print(tf.config.list_physical_devices('GPU'))
|
||||
|
||||
# Constants
|
||||
BATCH_SIZE = 8
|
||||
EPOCHS = 5
|
||||
EPOCHS = 50
|
||||
TRAIN_SAMPLES = 5
|
||||
|
||||
def load_list(list_path):
|
||||
|
|
@ -18,26 +18,18 @@ def load_list(list_path):
|
|||
video_details_list = json.load(json_file)
|
||||
return video_details_list
|
||||
|
||||
def load_frames_from_video(video_file, num_frames):
|
||||
print("Extracting video frames...")
|
||||
def load_frame_from_video(video_file):
|
||||
print("Extracting video frame...")
|
||||
cap = cv2.VideoCapture(video_file)
|
||||
frames = []
|
||||
count = 0
|
||||
while True:
|
||||
ret, frame = cap.read()
|
||||
if not ret:
|
||||
break
|
||||
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
||||
frames.append(frame)
|
||||
count += 1
|
||||
if count >= num_frames:
|
||||
break
|
||||
ret, frame = cap.read()
|
||||
if not ret:
|
||||
return None
|
||||
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
||||
cap.release()
|
||||
width, height = frame.shape[:2]
|
||||
return frames, width, height
|
||||
return frame
|
||||
|
||||
def preprocess(frames):
|
||||
return np.array(frames) / 255.0
|
||||
def preprocess(frame):
|
||||
return frame / 255.0
|
||||
|
||||
def save_model(model, file):
|
||||
os.makedirs("models", exist_ok=True)
|
||||
|
|
@ -54,109 +46,62 @@ def load_video_from_list(list_path):
|
|||
PRESET_SPEED = PRESET_SPEED_CATEGORIES.index(video_details['preset_speed'])
|
||||
video_details['preset_speed'] = PRESET_SPEED
|
||||
|
||||
train_frames, w, h = load_frames_from_video(os.path.join("test_data/", VIDEO_FILE), NUM_FRAMES * TRAIN_SAMPLES)
|
||||
frame = load_frame_from_video(os.path.join("test_data/", VIDEO_FILE))
|
||||
|
||||
all_frames.extend(train_frames)
|
||||
all_details.append({
|
||||
"frames": train_frames,
|
||||
"width": w,
|
||||
"height": h,
|
||||
"crf": CRF,
|
||||
"preset_speed": PRESET_SPEED,
|
||||
"video_file": VIDEO_FILE
|
||||
})
|
||||
if frame is not None:
|
||||
all_frames.append(preprocess(frame))
|
||||
all_details.append({
|
||||
"frame": frame,
|
||||
"crf": CRF,
|
||||
"preset_speed": PRESET_SPEED,
|
||||
"video_file": VIDEO_FILE
|
||||
})
|
||||
return all_details
|
||||
|
||||
def generate_frame_sequences(frames):
|
||||
sequences = []
|
||||
labels = []
|
||||
for i in range(len(frames) - NUM_FRAMES + 1):
|
||||
sequence = frames[i:i+NUM_FRAMES-1]
|
||||
sequences.append(sequence)
|
||||
labels.append(sequence[-1])
|
||||
return np.array(sequences), np.array(labels)
|
||||
|
||||
def frame_difference(frames):
|
||||
differences = []
|
||||
for i in range(1, len(frames)):
|
||||
differences.append(cv2.absdiff(frames[i], frames[i-1]))
|
||||
return differences
|
||||
|
||||
def main():
|
||||
all_video_details_train = load_video_from_list("test_data/training.json")
|
||||
all_video_details_val = load_video_from_list("test_data/validation.json")
|
||||
|
||||
model = VideoCompressionModel(NUM_CHANNELS, NUM_FRAMES)
|
||||
model = VideoCompressionModel(NUM_CHANNELS)
|
||||
model.compile(loss='mean_squared_error', optimizer='adam')
|
||||
early_stop = EarlyStopping(monitor='val_loss', patience=3, verbose=1, restore_best_weights=True)
|
||||
|
||||
# Load and concatenate all sequences and labels
|
||||
all_train_sequences = []
|
||||
all_val_sequences = []
|
||||
all_train_labels = []
|
||||
all_val_labels = []
|
||||
# Prepare data
|
||||
all_train_frames = []
|
||||
all_val_frames = []
|
||||
all_crf_train = []
|
||||
all_crf_val = []
|
||||
all_preset_speed_train = []
|
||||
all_preset_speed_val = []
|
||||
|
||||
for video_details_train, video_details_val in zip(all_video_details_train, all_video_details_val):
|
||||
train_frames = video_details_train["frames"]
|
||||
val_frames = video_details_val["frames"]
|
||||
|
||||
train_differences = frame_difference(preprocess(train_frames))
|
||||
val_differences = frame_difference(preprocess(val_frames))
|
||||
|
||||
#print(len(train_differences), train_differences[0].shape)
|
||||
|
||||
train_sequences, train_labels = generate_frame_sequences(train_differences)
|
||||
val_sequences, val_labels = generate_frame_sequences(val_differences)
|
||||
|
||||
crf_array_train = np.full((len(train_sequences), 1), video_details_train['crf'])
|
||||
crf_array_val = np.full((len(val_sequences), 1), video_details_val['crf'])
|
||||
preset_speed_array_train = np.full((len(train_sequences), 1), video_details_train['preset_speed'])
|
||||
preset_speed_array_val = np.full((len(val_sequences), 1), video_details_val['preset_speed'])
|
||||
|
||||
all_train_sequences.extend(train_sequences)
|
||||
all_val_sequences.extend(val_sequences)
|
||||
all_train_labels.extend(train_labels)
|
||||
all_val_labels.extend(val_labels)
|
||||
all_crf_train.extend(crf_array_train)
|
||||
all_crf_val.extend(crf_array_val)
|
||||
all_preset_speed_train.extend(preset_speed_array_train)
|
||||
all_preset_speed_val.extend(preset_speed_array_val)
|
||||
all_train_frames.append(video_details_train["frame"])
|
||||
all_val_frames.append(video_details_val["frame"])
|
||||
all_crf_train.append(video_details_train['crf'])
|
||||
all_crf_val.append(video_details_val['crf'])
|
||||
all_preset_speed_train.append(video_details_train['preset_speed'])
|
||||
all_preset_speed_val.append(video_details_val['preset_speed'])
|
||||
|
||||
# Convert lists to numpy arrays
|
||||
all_train_sequences = np.array(all_train_sequences)
|
||||
all_val_sequences = np.array(all_val_sequences)
|
||||
all_train_labels = np.array(all_train_labels)
|
||||
all_val_labels = np.array(all_val_labels)
|
||||
all_train_frames = np.array(all_train_frames)
|
||||
all_val_frames = np.array(all_val_frames)
|
||||
all_crf_train = np.array(all_crf_train)
|
||||
all_crf_val = np.array(all_crf_val)
|
||||
all_preset_speed_train = np.array(all_preset_speed_train)
|
||||
all_preset_speed_val = np.array(all_preset_speed_val)
|
||||
|
||||
# Shuffle the training data
|
||||
indices_train = np.arange(all_train_sequences.shape[0])
|
||||
np.random.shuffle(indices_train)
|
||||
|
||||
all_train_sequences = all_train_sequences[indices_train]
|
||||
all_train_labels = all_train_labels[indices_train]
|
||||
all_crf_train = all_crf_train[indices_train]
|
||||
all_preset_speed_train = all_preset_speed_train[indices_train]
|
||||
|
||||
print("\nTraining the model on mixed sequences...")
|
||||
print("\nTraining the model on frame pairs...")
|
||||
model.fit(
|
||||
{"frames": all_train_sequences, "crf": all_crf_train, "preset_speed": all_preset_speed_train},
|
||||
all_train_labels,
|
||||
{"frame": all_train_frames, "crf": all_crf_train, "preset_speed": all_preset_speed_train},
|
||||
all_val_frames, # Target is the compressed frame
|
||||
batch_size=BATCH_SIZE,
|
||||
epochs=EPOCHS,
|
||||
validation_data=({"frames": all_val_sequences, "crf": all_crf_val, "preset_speed": all_preset_speed_val}, all_val_labels),
|
||||
validation_data=({"frame": all_val_frames, "crf": all_crf_val, "preset_speed": all_preset_speed_val}, all_val_frames),
|
||||
callbacks=[early_stop]
|
||||
)
|
||||
print("\nTraining completed!")
|
||||
|
||||
save_model(model, 'model_differencing.keras')
|
||||
save_model(model, 'model.keras')
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
|
|
|
|||
Reference in a new issue