def extract_features(path_in, model_config, net, num_layers_finetune, use_gpu, num_timesteps=1, log_fn=print):
# Create inference engine
inference_engine = engine.InferenceEngine(net, use_gpu=use_gpu)
# extract features
for split in SPLITS:
videos_dir = directories.get_videos_dir(path_in, split)
features_dir = directories.get_features_dir(path_in, split, model_config, num_layers_finetune)
video_files = glob.glob(os.path.join(videos_dir, "*", "*.mp4"))
num_videos = len(video_files)
log_fn(f"\nFound {num_videos} videos to process in the {split}-set")
for video_index, video_path in enumerate(video_files):
log_fn(f'\rExtract features from video {video_index + 1} / {num_videos}')
path_features = video_path.replace(videos_dir, features_dir).replace(".mp4", ".npy")
if os.path.isfile(path_features):
log_fn("\tSkipped - feature was already precomputed.")
else:
# Read all frames
frames = extract_frames(video_path=video_path,
inference_engine=inference_engine)
compute_features(path_features=path_features,
inference_engine=inference_engine,
frames=frames,
batch_size=16,
num_timesteps=num_timesteps)
log_fn('\n')
def extract_features(path_in, net, num_layers_finetune, use_gpu, num_timesteps=1):
# Create inference engine
inference_engine = engine.InferenceEngine(net, use_gpu=use_gpu)
# extract features
for dataset in ["train", "valid"]:
videos_dir = os.path.join(path_in, f"videos_{dataset}")
features_dir = os.path.join(path_in, f"features_{dataset}_num_layers_to_finetune={num_layers_finetune}")
video_files = glob.glob(os.path.join(videos_dir, "*", "*.mp4"))
print(f"\nFound {len(video_files)} videos to process in the {dataset}set")
for video_index, video_path in enumerate(video_files):
print(f"\rExtract features from video {video_index + 1} / {len(video_files)}",
end="")
path_out = video_path.replace(videos_dir, features_dir).replace(".mp4", ".npy")
if os.path.isfile(path_out):
print("\n\tSkipped - feature was already precomputed.")
else:
# Read all frames
compute_features(video_path, path_out, inference_engine,
num_timesteps=num_timesteps, path_frames=None, batch_size=16)
print('\n')
def load_feature_extractor(project_path):
feature_extractor = backbone_networks.StridedInflatedEfficientNet()
# Remove internal padding for feature extraction and training
checkpoint = torch.load('resources/backbone/strided_inflated_efficientnet.ckpt')
feature_extractor.load_state_dict(checkpoint)
feature_extractor.eval()
# Create Inference Engine
use_gpu = get_project_setting(project_path, 'use_gpu')
inference_engine = engine.InferenceEngine(feature_extractor, use_gpu=use_gpu)
return inference_engine
def _load_feature_extractor():
global inference_engine
import torch
from sense import engine
from sense import feature_extractors
if inference_engine is None:
feature_extractor = feature_extractors.StridedInflatedEfficientNet()
# Remove internal padding for feature extraction and training
checkpoint = torch.load('resources/backbone/strided_inflated_efficientnet.ckpt')
feature_extractor.load_state_dict(checkpoint)
feature_extractor.eval()
# Create Inference Engine
inference_engine = engine.InferenceEngine(feature_extractor, use_gpu=True)
if __name__ == "__main__":
# Parse argument
args = docopt(__doc__)
dataset_path = join(os.getcwd(), args['--data_path'])
# Load feature extractor
feature_extractor = feature_extractors.StridedInflatedEfficientNet()
# Remove internal padding for feature extraction and training
checkpoint = torch.load(
'resources/backbone/strided_inflated_efficientnet.ckpt')
feature_extractor.load_state_dict(checkpoint)
feature_extractor.eval()
# Create Inference Engine
inference_engine = engine.InferenceEngine(feature_extractor, use_gpu=True)
for split in ['train', 'valid']:
print("\n" + "-" * 10 + f"Preparing videos in the {split}-set" +
"-" * 10)
for label in os.listdir(join(dataset_path, f'videos_{split}')):
# Get data-set from path, given split and label
folder = join(dataset_path, f'videos_{split}', label)
# Create features and frames folders for the given split and label
features_folder = join(dataset_path, f'features_{split}', label)
frames_folder = join(dataset_path, f'frames_{split}', label)
os.makedirs(features_folder, exist_ok=True)
os.makedirs(frames_folder, exist_ok=True)
# Loop through all videos for the given class-label