def __init__(self, model_path, with_cuda, yolact_config, fast_nms,
threshold, display_cv, top_k):
self.top_k = top_k
self.threshold = threshold
self.display_cv = display_cv
print("loading Yolact ...")
with torch.no_grad():
set_cfg(yolact_config)
print("Configuration: ", yolact_config)
if with_cuda:
cudnn.benchmark = True
cudnn.fastest = True
torch.set_default_tensor_type('torch.cuda.FloatTensor')
else:
torch.set_default_tensor_type('torch.FloatTensor')
print("use cuda: ", with_cuda)
self.net = Yolact()
self.net.load_weights(model_path)
print("Model: ", model_path)
self.net.eval()
if with_cuda:
self.net = self.net.cuda()
self.net.detect.use_fast_nms = fast_nms
print("use fast nms: ", fast_nms)
print("Yolact loaded")
def main():
parse_args()
rospy.init_node('yolact_ros', anonymous=True)
if args.config is not None:
set_cfg(args.config)
if args.config is None:
model_path = SavePath.from_str(args.trained_model)
# TODO: Bad practice? Probably want to do a name lookup instead.
args.config = model_path.model_name + '_config'
print('Config not specified. Parsed %s from the file name.\n' %
args.config)
set_cfg(args.config)
if args.detect:
cfg.eval_mask_branch = False
if args.dataset is not None:
set_dataset(args.dataset)
with torch.no_grad():
if not os.path.exists('results'):
os.makedirs('results')
if args.cuda:
cudnn.benchmark = True
cudnn.fastest = True
torch.set_default_tensor_type('torch.cuda.FloatTensor')
else:
torch.set_default_tensor_type('torch.FloatTensor')
if args.resume and not args.display:
with open(args.ap_data_file, 'rb') as f:
ap_data = pickle.load(f)
calc_map(ap_data)
exit()
print('Loading model...', end='')
net = Yolact()
net.load_weights(args.trained_model)
net.eval()
print(' Done.')
if args.cuda:
net = net.cuda()
net.detect.use_fast_nms = True
cfg.mask_proto_debug = False
detect_ = DetectImg(net)
try:
rospy.spin()
except KeyboardInterrupt:
print("Shutting down")
cv2.destroyAllWindows()
#pub = rospy.Publisher('chatter',String,queue_size=10)
#rate = rospy.Rate(50) #10hz
#str_ += text_str
#rospy.loginfo(str_)
#pub.publish(str_)
#rate.sleep()
return img_numpy
if __name__ == '__main__':
parse_args()
if args.config is not None:
set_cfg(args.config)
if args.trained_model == 'interrupt':
args.trained_model = SavePath.get_interrupt('weights/')
elif args.trained_model == 'latest':
args.trained_model = SavePath.get_latest('weights/', cfg.name)
if args.config is None:
model_path = SavePath.from_str(args.trained_model)
# TODO: Bad practice? Probably want to do a name lookup instead.
args.config = model_path.model_name + '_config'
print('Config not specified. Parsed %s from the file name.\n' %
args.config)
set_cfg(args.config)
if args.detect:
else:
pred_outs['conf'] = F.softmax(pred_outs['conf'], -1)
return self.detect(pred_outs, self)
# Some testing code
if __name__ == '__main__':
from yolact.utils.functions import init_console
init_console()
# Use the first argument to set the config if you want
import sys
if len(sys.argv) > 1:
from yolact.data.config import set_cfg
set_cfg(sys.argv[1])
net = Yolact()
net.train()
net.init_weights(backbone_path='weights/' + cfg.backbone.path)
# GPU
net = net.cuda()
torch.set_default_tensor_type('torch.cuda.FloatTensor')
x = torch.zeros((1, 3, cfg.max_size, cfg.max_size))
y = net(x)
for p in net.prediction_layers:
print(p.last_conv_size)