def compare():
date_time_now = str(datetime.datetime.now()).replace(" ", "_").replace(
":", "_")
compare_path = f"../5Compare/individual_train/{date_time_now}/"
if not os.path.exists(compare_path):
os.makedirs(compare_path)
config_name = "exp_config.p"
conf_list = np.arange(start=0.1, stop=0.95, step=0.025)
seed_range = range(420, 425)
classes = load_classes('../4Others/color_ball.names')
cfg_path = "../4Others/color_ball.cfg"
blocks = parse_cfg(cfg_path)
model = yolo_v3(blocks)
model.load_weights("../4Weights/yolov3.weights", cust_train_zero=True)
with open(compare_path + config_name, "wb") as fp:
pickle.dump(model.net, fp, protocol=pickle.HIGHEST_PROTOCOL)
time_taken_df = pd.DataFrame(columns=list(seed_range))
for cls_index, cls in enumerate(classes):
to_path = f"../{cls}/"
for index, seed in enumerate(seed_range):
model.load_weights("../4Weights/yolov3.weights",
cust_train_zero=True)
x_start = time.time()
random.seed(seed)
if not os.path.exists(to_path):
os.makedirs(to_path)
sub_name = f"{cls}_seed_{seed}_"
name_list = ["img_name", "c", "gx", "gy", "gw", "gh"]
# Original label names
label_csv_mame = '../color_balls/label.csv'
img_txt_path = "../color_balls/*.txt"
prep_labels(img_txt_path, name_list, label_csv_mame)
# sub sampled label names
sub_sample_csv_name = to_path + "label.csv"
sub_sample_txt_path = to_path + "*.txt"
prep_labels(sub_sample_txt_path, name_list, sub_sample_csv_name)
# label_csv_mame = '../1TestData/label.csv'
# img_txt_path = "../1TestData/*.txt"
move_images(label_name=label_csv_mame,
to_path=to_path,
action_fn=action_fn,
cat_nums=[cls_index])
best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap,\
map_frame = main(model,
classes, conf_list,
sub_sample_csv_name,
sub_sample_txt_path,
to_path,
cuda=True,
specific_conf=0.5,
sub_name=sub_name,
selected_cls=[str(x) for x in [cls_index]])
map_frame.to_csv(f"{compare_path+sub_name}.csv", index=True)
shutil.rmtree(to_path)
time_taken_df.loc[cls, seed] = time.time() - x_start
# if you change the csv_name, pls change accordingly in the
# visualization part
time_taken_df.to_csv(compare_path + 'time_taken.csv', index=True)
def test_loaders():
label_csv_mame = '../1TestData/label.csv'
name_list = ["img_name", "c", "gx", "gy", "gw", "gh"]
test_img_txt_path = "../1TestData/*.txt"
prep_labels(test_img_txt_path, name_list, label_csv_mame)
logging.basicConfig(level=logging.DEBUG,
format="[%(asctime)s %(filename)s] %(message)s")
# training transformation
pre_trans = RandomCrop(jitter=0.25,
inp_dim=480)
train_transform = transforms.Compose(
[transforms.ColorJitter(
brightness=0.25,
saturation=0.25
),
transforms.ToTensor()
])
test_pre_trans = OfssetLabels(resize=True, input_dim=480)
test_transform = transforms.Compose(
[ImgToTensorCv()])
train_data = CustData('../1TrainData/label.csv',
pre_trans=pre_trans,
transform=train_transform)
# test_data = CustData(test_label_csv_mame,
# transform=test_transform)
test_data = CustDataCV('../1TestData/label.csv',
pre_trans=test_pre_trans,
transform=test_transform)
train_loader = DataLoader(train_data, shuffle=False,
batch_size=2,
collate_fn=my_collate,
num_workers=0,
worker_init_fn=worker_init_fn)
test_loader = DataLoader(test_data, shuffle=False,
batch_size=4,
collate_fn=my_collate,
num_workers=0,
worker_init_fn=worker_init_fn)
# print("running train loader")
# for step, samples in enumerate(train_loader):
# if step < 200:
# images, labels = samples["image"], samples["label"]
# for img, label in zip(images, labels):
# img = img.permute(1, 2, 0).contiguous().numpy()
# draw_boxes(img, label)
# else:
# break
print("running test loader")
for step, samples in enumerate(test_loader):
if step < 2:
images, labels = samples["image"], samples["label"]
for img, label in zip(images, labels):
img = img.permute(1, 2, 0).contiguous().numpy()
draw_boxes(img, label)
else:
break
def cross_validation():
config = {'label_csv_mame': '../2CvTrain/label.csv',
'test_label_csv_mame': '../1TestData/label.csv',
'test_img_txt_path': "../1TestData/*.txt",
'valid_label_csv_mame': '../2CvValid/label.csv',
'valid_img_txt_path': "../2CvValid/*.txt",
# label csv column names
'name_list': ["img_name", "c", "gx", "gy", "gw", "gh"],
'cfg_path': "../4Others/yolo.cfg",
'params_dir': '../4Others/params.txt'}
# entire trainin label
prep_label_config = {'label_csv_mame': '../1TrainData/label.csv',
'img_txt_path': "../1TrainData/*.txt",
'name_list': config['name_list']}
# cv train label
cv_train_label_config = {'label_csv_mame': config['label_csv_mame'],
'img_txt_path': "../2CvTrain/*.txt",
'name_list': config['name_list']}
torch.backends.cudnn.enabled = False
torch.backends.cudnn.benchmark = False
# you can specify any tuning hyper-parammeters here , but do not use
# np.arrange() as it will result an error for json.dumps()
tune_params = {'seed': list(range(1, 5)),
'batch_size': list(range(2, 7, 2)),
'height': [416, 448]
}
index, values = zip(*tune_params.items())
experiments_params = [dict(zip(index, v)) for v in it.product(*values)]
# prepare label for the entire training set
prep_labels(**prep_label_config)
cv_results_list = []
for experiment_params in experiments_params:
cv_split_config = {'n_splits': 5,
'cv': True,
'train_size': 0.7,
'name_list': config['name_list'],
'random_state': experiment_params['seed'],
'train_cv_path': "../2CvTrain/",
'valid_cv_path': "../2CvValid/",
'label_name': prep_label_config['label_csv_mame']}
for _ in split_train(**cv_split_config):
prep_labels(**cv_train_label_config)
params = prep_params(config['params_dir'],
cv_train_label_config['label_csv_mame'],
experiment_params)
best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap,\
map_frame = run_training(params=params, **config)
cv_results_list.append(params)
shutil.rmtree(cv_split_config['train_cv_path'])
shutil.rmtree(cv_split_config['valid_cv_path'])
cv_results_df = pd.DataFrame(cv_results_list)
cv_results_df.to_csv(f"{params['working_dir']}cv_results.csv")
def main(weight_path):
# change cuda to True if you wish to use gpu
det = "../2ProcessedData/"
cfg_path = "../4Others/yolo.cfg"
label_csv_mame = '../1TestData/label.csv'
name_list = ["img_name", "c", "gx", "gy", "gw", "gh"]
test_img_txt_path = "../1TestData/*.txt"
prep_labels(test_img_txt_path, name_list, label_csv_mame)
params = prep_params(weight_path, label_csv_mame)
detection(cfg_path, det, label_csv_mame, params)
df = get_single_img_map(cfg_path, det, label_csv_mame, params)
return df
'label_csv_mame': config['label_csv_mame'],
'img_txt_path': "../1TrainData/*.txt",
'name_list': config['name_list']
}
# torch.backends.cudnn.enabled = False
# torch.backends.cudnn.benchmark = False
# you can specify any tuning hyper-parammeters here , but do not use
# np.arrange() as it will result an error for json.dumps()
tune_params = {
'seed': list(range(1, 3)),
'height': [448, 480],
"decay": [0.0005, 0.001],
"steps": [6, 10],
"optimizer": ['sgd', 'adam'],
"epochs": [40, 80]
}
index, values = zip(*tune_params.items())
experiments_params = [dict(zip(index, v)) for v in it.product(*values)]
# prepare training lables
prep_labels(**prep_label_config)
for experiment_params in experiments_params:
params = prep_params(config['params_dir'], config['label_csv_mame'],
experiment_params)
try:
best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap,\
map_frame = run_training(params=params, **config)
except:
continue
def compare():
date_time_now = str(datetime.datetime.now()).replace(" ", "_").replace(
":", "_")
compare_path = f"../5Compare/epoch_experiment/{date_time_now}/"
if not os.path.exists(compare_path):
os.makedirs(compare_path)
config_name = "exp_config.p"
conf_list = np.arange(start=0.2, stop=0.95, step=0.025)
seed_range = range(420, 425)
to_path_list = {
"../250_to_300_imgs/": 1.05 / 7,
"../400_to_450_imgs/": 1 / 4,
"../550_to_600_imgs/": 3 / 7,
"../700_to_750_imgs/": 4.7 / 7
}
epochs_list = [25, 30, 35, 40]
classes = load_classes('../4Others/color_ball.names')
cfg_path = "../4Others/color_ball.cfg"
blocks = parse_cfg(cfg_path)
model = yolo_v3(blocks)
model.load_weights("../4Weights/yolov3.weights", cust_train_zero=True)
model.net['epochs_list'] = epochs_list
model.net['img_sampling_info'] = to_path_list
with open(compare_path + config_name, "wb") as fp:
pickle.dump(model.net, fp, protocol=pickle.HIGHEST_PROTOCOL)
for to_path, train_size in to_path_list.items():
time_taken_df = pd.DataFrame(columns=list(epochs_list))
file_name = to_path.strip(".").strip("/")
for index, seed in enumerate(seed_range):
x_start = time.time()
random.seed(seed)
if not os.path.exists(to_path):
os.makedirs(to_path)
name_list = ["img_name", "c", "gx", "gy", "gw", "gh"]
# Original label names
label_csv_mame = '../color_balls/label.csv'
img_txt_path = "../color_balls/*.txt"
prep_labels(img_txt_path, name_list, label_csv_mame)
# sub sampled label names
sub_sample_csv_name = to_path + "label.csv"
sub_sample_txt_path = to_path + "*.txt"
# label_csv_mame = '../1TestData/label.csv'
# img_txt_path = "../1TestData/*.txt"
move_images(label_name=label_csv_mame,
to_path=to_path,
action_fn=split_train,
train_size=train_size,
random_state=seed)
for epoch in epochs_list:
model.load_weights("../4Weights/yolov3.weights",
cust_train_zero=True)
model.net['epochs'] = epoch
sub_name = f"{file_name}_seed_{seed}_epoch_{epoch}_"
best_map, best_ap, best_conf, specific_conf_map,\
specific_conf_ap, map_frame = main(model,
classes, conf_list,
sub_sample_csv_name,
sub_sample_txt_path,
to_path,
cuda=True,
specific_conf=0.5,
sub_name=sub_name)
map_frame.to_csv(f"{compare_path+sub_name}.csv", index=True)
time_taken_df.loc[file_name, epoch] = time.time() - x_start
# if you change the csv_name, pls change accordingly in the
# visualization part
time_taken_df.to_csv(compare_path + f"time_taken_{seed}.csv",
index=True)
shutil.rmtree(to_path)
def run_training(params,
label_csv_mame,
name_list,
test_label_csv_mame,
test_img_txt_path,
cfg_path,
params_dir,
valid_label_csv_mame=False,
valid_img_txt_path=False):
random.seed(params['seed'])
np.random.seed(params['seed'])
torch.manual_seed(params['seed'])
torch.cuda.manual_seed(params['seed'])
logging.basicConfig(level=logging.DEBUG,
format="[%(asctime)s %(filename)s] %(message)s")
# prepare test lables
prep_labels(test_img_txt_path, name_list, test_label_csv_mame)
# parse model architect from architect config file
blocks = parse_cfg(cfg_path)
# training transformation
pre_trans = RandomCrop(jitter=params['rand_crop'],
inp_dim=params["height"])
train_transform = transforms.Compose([
transforms.ColorJitter(brightness=params["exposure"],
saturation=params["saturation"]),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])
test_pre_trans = OfssetLabels(resize=True, input_dim=params['height'])
test_transform = transforms.Compose([
ImgToTensorCv(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])
train_data = CustData(label_csv_mame,
pre_trans=pre_trans,
transform=train_transform)
test_data = CustDataCV(test_label_csv_mame,
pre_trans=test_pre_trans,
transform=test_transform)
train_loader = DataLoader(train_data,
shuffle=True,
batch_size=params["batch_size"],
collate_fn=my_collate,
num_workers=params['num_workers'],
worker_init_fn=worker_init_fn)
test_loader = DataLoader(test_data,
shuffle=False,
batch_size=params["batch_size"],
collate_fn=my_collate,
num_workers=params['num_workers'],
worker_init_fn=worker_init_fn)
# initiate model
model = yolo_v3(params, blocks)
# Initiate validation data loader if there is any
if valid_label_csv_mame:
valid_data = CustDataCV(valid_label_csv_mame, transform=test_transform)
valid_loader = DataLoader(valid_data,
shuffle=False,
batch_size=params["batch_size"],
collate_fn=my_collate,
num_workers=params['num_workers'],
worker_init_fn=worker_init_fn)
# Start training
best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap,\
map_frame = model.fit(train_loader, valid_loader, test_loader)
else:
# Start training
best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap, \
map_frame = model.fit(train_loader, test_loader)
return best_map, best_ap, best_conf, specific_conf_map, specific_conf_ap, \
map_frame