def main():
for epoch in parse_epoch(args.epoch):
# get checkpoint
checkpoint_pth = get_checkpoint_path(args.training_base_param, epoch)
if args.exp_id == '':
exp_id = args.training_base_param + '.epoch{}'.format(epoch)
else:
exp_id = args.exp_id
print('=> Evaluating: {}'.format(exp_id))
if args.dataset in ['CVPR13', 'OTB50', 'OTB100']:
# for OTB datasets, we save results into the same directory
save_dataset_name = 'OTB100'
else:
save_dataset_name = args.dataset
# set up parameters
params = {
'dataset_name': args.dataset,
'checkpoint': checkpoint_pth,
'tracking_base_param': args.tracking_base_param,
'num_repeat': args.num_repeat,
'exp_id': exp_id,
'result_dir': env_settings().results_path,
'save_dataset_name': save_dataset_name,
}
metrics = run_tracking_and_evaluate(params)
save_info(params, metrics)
def initialize(self):
with fluid.dygraph.guard():
if os.path.isabs(self.net_path):
net_path_full = self.net_path
else:
net_path_full = os.path.join(env_settings().network_path,
self.net_path)
self.net = siamfc_alexnet(
backbone_pretrained=False,
backbone_is_test=True,
estimator_is_test=True)
state_dictsm, _ = fluid.load_dygraph(net_path_full)
self.net.load_dict(state_dictsm)
self.net.train()
self.target_estimator = self.net.target_estimator
self.layer_stride = {'conv5': 8}
self.layer_dim = {'conv5': 256}
self.estimator_feature_layers = self.net.target_estimator_layer
if isinstance(self.pool_stride, int) and self.pool_stride == 1:
self.pool_stride = [1] * len(self.output_layers)
self.feature_layers = sorted(
list(set(self.output_layers + self.estimator_feature_layers)))
self.mean = np.reshape([0., 0., 0.], [1, -1, 1, 1])
self.std = np.reshape([1 / 255., 1 / 255., 1 / 255.], [1, -1, 1, 1])
def run_tracking_and_evaluate(params):
"""Receive hyperparameters and return the evaluation metric"""
# load dataset
root = os.path.abspath(
osp.join(env_settings().dataset_path, params['save_dataset_name']))
dataset = DatasetFactory.create_dataset(name=params['dataset_name'],
dataset_root=root)
run_one_dataset(dataset, params)
metrics = compute_evaluation_metrics(dataset, params)
return metrics
def initialize(self):
with fluid.dygraph.guard():
if os.path.isabs(self.net_path):
net_path_full = self.net_path
else:
net_path_full = os.path.join(env_settings().network_path, self.net_path)
self.net = SiamMask_ResNet50_sharp(backbone_pretrained=False, is_test=True)
state_dict, _ = fluid.load_dygraph(net_path_full)
self.net.load_dict(state_dict)
self.net.eval()
def initialize(self):
with fluid.dygraph.guard():
if os.path.isabs(self.net_path):
net_path_full = self.net_path
else:
net_path_full = os.path.join(env_settings().network_path,
self.net_path)
self.net = atom_resnet18(
backbone_pretrained=False,
backbone_is_test=True,
iounet_is_test=True)
state_dictsm, _ = fluid.load_dygraph(net_path_full)
self.net.load_dict(state_dictsm)
self.net.train()
self.iou_predictor = self.net.bb_regressor
self.layer_stride = {
'conv0': 2,
'conv1': 2,
'block0': 4,
'block1': 8,
'block2': 16,
'block3': 32,
'classification': 16,
'fc': None
}
self.layer_dim = {
'conv0': 64,
'conv1': 64,
'block0': 64,
'block1': 128,
'block2': 256,
'block3': 512,
'classification': 256,
'fc': None
}
self.iounet_feature_layers = self.net.bb_regressor_layer
if isinstance(self.pool_stride, int) and self.pool_stride == 1:
self.pool_stride = [1] * len(self.output_layers)
self.feature_layers = sorted(
list(set(self.output_layers + self.iounet_feature_layers)))
self.mean = np.reshape([0.485, 0.456, 0.406], [1, -1, 1, 1])
self.std = np.reshape([0.229, 0.224, 0.225], [1, -1, 1, 1])
def get_checkpoint_path(training_base_param, epoch):
model_dir = osp.abspath(
osp.join(env_settings().network_path, *training_base_param.split('.')))
model_names = glob(model_dir + '/*.pdparams')
prefix = '_'.join(model_names[0].split('_')[:-1])
return osp.join(model_dir, '{}_ep{:04d}'.format(prefix, epoch))