def main():
args = parse_args()
net = CrowdCounter(is_cuda)
network.load_net(args.model_path, net)
if is_cuda:
net.cuda()
print('eval:', net.eval())
video = cv2.VideoCapture(args.video)
nframe = 0
hist = []
while 1:
ok, frame = video.read()
frame = cv2.resize(frame, (0, 0), fx=args.resize_fx, fy=args.resize_fx)
nframe += 1
if nframe % args.estimate_rate != 0:
continue
print('nframe', nframe)
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
imgarr = image_to_array(gray)
den, cnt = run(net, imgarr)
"""
if len(hist) > 5:
hist.pop(0)
hist.append(den)
histden = sum(hist) / len(hist)
"""
if show(frame, den, cnt):
break
def evaluate_model(trained_model, data_loader):
net = CrowdCounter()
network.load_net(trained_model, net)
net.cuda()
net.eval()
mae = 0.0
mse = 0.0
for blob in data_loader:
im_data = blob['data']
gt_data = blob['gt_density']
density_map = net(im_data, gt_data)
density_map = density_map.data.cpu().numpy()
gt_count = np.sum(gt_data)
et_count = np.sum(density_map)
mae += abs(gt_count - et_count)
mse += ((gt_count - et_count) * (gt_count - et_count))
mae = mae / data_loader.get_num_samples()
mse = np.sqrt(mse / data_loader.get_num_samples())
return mae, mse
def main():
args = parse_args()
net = CrowdCounter(is_cuda)
network.load_net(args.model_path, net)
if is_cuda:
net.cuda()
print('eval:', net.eval())
for filename in args.image_files:
img = cv2.imread(filename)
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
imgarr = image_to_array(gray)
den, cnt = run(net, imgarr)
#if show(img, den, cnt): break
display_results_count(img, den, cnt)
output_dir = './output/'
model_name = os.path.basename(model_path).split('.')[0]
file_results = os.path.join(output_dir,'results_' + model_name + '_.txt')
if not os.path.exists(output_dir):
os.mkdir(output_dir)
output_dir = os.path.join(output_dir, 'density_maps_' + model_name)
if not os.path.exists(output_dir):
os.mkdir(output_dir)
net = CrowdCounter()
trained_model = os.path.join(model_path)
network.load_net(trained_model, net)
net.cuda()
net.eval()
mae = 0.0
mse = 0.0
#load test data
data_loader = ImageDataLoader(data_path, gt_path, shuffle=False, gt_downsample=True, pre_load=True)
for blob in data_loader:
im_data = blob['data']
gt_data = blob['gt_density']
density_map = net(im_data, gt_data)
density_map = density_map.data.cpu().numpy()
gt_count = np.sum(gt_data)
et_count = np.sum(density_map)
mae += abs(gt_count-et_count)
mse += ((gt_count-et_count)*(gt_count-et_count))
def main():
parser = argparse.ArgumentParser(description='mcnn worldexp.')
parser.add_argument('--preload', type=int, default=0)
parser.add_argument('--data',
type=str,
default="./minibus_data",
help='data path')
parser.add_argument('--model',
type=str,
default="./saved_models/mcnn_worldexpo_2000.h5",
help='model path')
parser.add_argument('--saveden', type=int, default=1)
parser.add_argument('--vis', type=int, default=0)
args = parser.parse_args()
method = 'mcnn'
dataset_name = 'minibus'
output_dir = './saved_models/'
data_path = args.data
vis = args.vis
gt_path = None
model_path = args.model
save_output = args.saveden
output_dir = './output/'
model_name = os.path.basename(model_path).split('.')[0]
file_results = os.path.join(output_dir, 'results_' + model_name + '_.txt')
if not os.path.exists(output_dir):
os.mkdir(output_dir)
output_dir = os.path.join(output_dir, 'density_maps_' + model_name)
if not os.path.exists(output_dir):
os.mkdir(output_dir)
net = CrowdCounter()
trained_model = os.path.join(model_path)
network.load_net(trained_model, net)
print("loaded: {}".format(trained_model))
net.cuda()
net.eval()
mae = 0.0
mse = 0.0
#load test data
data_loader = ExrImageDataLoader(data_path,
None,
mask_path=-1,
shuffle=False,
gt_downsample=True,
pre_load=args.preload)
for blob in data_loader:
im_data = blob['data']
gt_data = blob['gt_density']
mask = blob["mask"]
density_map = net(im_data, gt_data, mask)
density_map = density_map.data.cpu().numpy()
gt_count = np.asscalar(np.sum(gt_data))
et_count = np.asscalar(np.sum(density_map))
log_text = 'fname: %s gt_cnt: %4.1f, et_cnt: %4.1f' % (
blob['fname'], gt_count, et_count)
print(log_text)
mae += abs(gt_count - et_count)
mse += ((gt_count - et_count) * (gt_count - et_count))
if vis:
utils.display_results(im_data, gt_data, density_map)
if save_output:
utils.save_density_map(
density_map, output_dir,
'output_' + blob['fname'].split('.')[0] + '.png')
mae = mae / data_loader.get_num_samples()
mse = np.sqrt(mse / data_loader.get_num_samples())
print('\nMAE: %0.2f, MSE: %0.2f' % (mae, mse))