Spaces:
Running
Running
| from __future__ import annotations | |
| from typing import List | |
| import cv2 | |
| import os | |
| import tensorflow as tf | |
| # Disable all GPUS | |
| tf.config.set_visible_devices([], 'GPU') | |
| vocab = [x for x in "abcdefghijklmnopqrstuvwxyz'?!123456789 "] | |
| char_to_num = tf.keras.layers.StringLookup(vocabulary=vocab, oov_token="") | |
| num_to_char = tf.keras.layers.StringLookup( | |
| vocabulary=char_to_num.get_vocabulary(), oov_token="", invert=True | |
| ) | |
| def load_video(path: str) -> List[float]: | |
| cap = cv2.VideoCapture(path) | |
| frames = [] | |
| for _ in range(int(cap.get(cv2.CAP_PROP_FRAME_COUNT))): | |
| ret, frame = cap.read() | |
| if not ret or frame is None: | |
| break | |
| frame = tf.image.rgb_to_grayscale(tf.cast(frame, tf.float32)) | |
| frames.append(frame[190:236, 80:220, :]) | |
| cap.release() | |
| if not frames: | |
| raise ValueError(f"No frames were read from video: {path}") | |
| mean = tf.math.reduce_mean(frames) | |
| std = tf.math.reduce_std(tf.cast(frames, tf.float32)) | |
| return tf.cast((frames - mean), tf.float32) / std | |
| def load_alignments(path: str) -> List[str]: | |
| with open(path, 'r') as f: | |
| lines = f.readlines() | |
| tokens = [] | |
| for line in lines: | |
| line = line.split() | |
| if len(line) < 3: | |
| continue | |
| if line[2] != 'sil': | |
| tokens = [*tokens, ' ', line[2]] | |
| return char_to_num( | |
| tf.reshape(tf.strings.unicode_split(tokens, input_encoding='UTF-8'), (-1)) | |
| )[1:] | |
| def load_data(path: str): | |
| path = bytes.decode(path.numpy()) | |
| file_name = os.path.splitext(os.path.basename(path))[0] | |
| BASE_DIR = os.path.dirname(os.path.abspath(__file__)) | |
| data_dir = os.path.abspath(os.path.join(BASE_DIR, 'data', 's1')) | |
| alignment_dir = os.path.abspath(os.path.join(BASE_DIR, 'data', 'alignments', 's1')) | |
| video_path = os.path.join(data_dir, f'{file_name}.mpg') | |
| alignment_path = os.path.join(alignment_dir, f'{file_name}.align') | |
| if not os.path.exists(video_path): | |
| raise FileNotFoundError(f"Video file {video_path} does not exist.") | |
| if not os.path.exists(alignment_path): | |
| raise FileNotFoundError(f"Alignment file {alignment_path} does not exist.") | |
| frames = load_video(video_path) | |
| alignments = load_alignments(alignment_path) | |
| return frames, alignments | |