def main(): args = parse_args() logger = log.get_logger(args.log) args.logger = logger logger.info('*' * 80) logger.info('the args are the below') logger.info('*' * 80) for x in args.__dict__: logger.info(x + ',' + str(args.__dict__[x])) logger.info(cfg.config[args.dataset]) logger.info('*' * 80) os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu if not os.path.exists(args.param_dir): os.mkdir(args.param_dir) torch.manual_seed(long(time.time())) model = bdcn.BDCN(pretrain=args.pretrain, logger=logger) if args.complete_pretrain: model.load_state_dict(torch.load(args.complete_pretrain)) logger.info(model) train(model, args)
def main(): args = parse_args() # Choose the GPUs os.environ["CUDA_DEVICE_ORDER"] = 'PCI_BUS_ID' os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu logger = log.get_logger(args.log) args.logger = logger logger.info('*' * 80) logger.info('the args are the below') logger.info('*' * 80) for x in args.__dict__: logger.info(x + ',' + str(args.__dict__[x])) logger.info('*' * 80) if not os.path.exists(args.param_dir): os.mkdir(args.param_dir) torch.manual_seed(int(time.time())) model = bdcn.BDCN(pretrain=args.pretrain, logger=logger) if args.complete_pretrain: model.load_state_dict(torch.load(args.complete_pretrain)) train(model, args)
def main(): import time print(time.localtime()) args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu model = bdcn.BDCN() model.load_state_dict(torch.load('%s' % (args.model))) test(model, args)
def main(): import time print time.localtime() args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu model = bdcn.BDCN(rate=args.rate) model.load_state_dict(torch.load('%s' % (args.model))) # print model.fuse.weight.data, model.fuse.bias.data print model.fuse.weight.data test(model, args)
def main(): import time print (time.localtime()) args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu device = torch.device('cpu' if torch.cuda.device_count() == 0 else 'cuda') model = bdcn.BDCN().to(device) model.load_state_dict(torch.load('%s' % (args.model), map_location=device)) test(model, args, running_on=device)
def main(): import time print(time.localtime()) args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu model = bdcn.BDCN() model.load_state_dict( torch.load('%s' % (params_dir[args.train_dataset]), map_location=torch.device('cpu'))) test(model, args)
def setup(opts): model = bdcn.BDCN() if torch.cuda.is_available(): model.load_state_dict(torch.load(opts["checkpoint"])) model.cuda() else: model.load_state_dict( torch.load(opts["checkpoint"], map_location='cpu')) model.eval() return model
def main(): import time print(time.localtime()) args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu device = torch.device('cpu' if torch.cuda.device_count() == 0 else 'cuda') # model = bdcn.BDCN() model = bdcn.BDCN().to(device) model_dir = os.path.join('params', args.train_data, args.model) model.load_state_dict(torch.load('%s' % (model_dir), map_location=device)) print("====== Checkpoint> ", model_dir, "==========") test(model, args, running_on=device)
def main(): inputDir = ['images'] args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu logging.info('Loading model...') model = bdcn.BDCN() logging.info('Loading state...') model.load_state_dict(torch.load('%s' % (args.model))) logging.info('Start image processing...') for inputDir in inputDir: args.inputDir = inputDir args.cuda = True forwardAll(model, args)
def main(): args = parse_args() # Choose the GPUs os.environ["CUDA_DEVICE_ORDER"] = 'PCI_BUS_ID' os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu model = bdcn.BDCN() if torch.cuda.is_available(): map_location = lambda storage, loc: storage.cuda() else: map_location = 'cpu' model.load_state_dict( torch.load('%s' % (args.model), map_location=map_location)) test(model, args)
def execute(gpu, cuda, trained_model, res_dir, data_root, test_lst): # Import model and use gpu os.environ['CUDA_VISIBLE_DEVICES'] = gpu model = bdcn.BDCN() ##### Using GPU #### if cuda: model.load_state_dict(torch.load('%s' % (trained_model))) else: #### Using CPU #### model.load_state_dict( torch.load('%s' % (trained_model), map_location=torch.device('cpu'))) # Call test function and detect edges test(model, cuda, res_dir, data_root, test_lst)
def main(): args = parse_args() os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu logging.info('Loading model...') model = bdcn.BDCN() logging.info('Loading state...') model.load_state_dict(torch.load('%s' % (args.model))) logging.info('Start image processing...') inputDirs = [ '/run/user/1000/gvfs/smb-share:server=192.168.0.253,share=data/Master/datasets/bsr_bsds500/BSR/BSDS500/data/images/test_png/' #'/run/user/1000/gvfs/smb-share:server=192.168.0.253,share=data/Master/datasets/test_dataset/hdr_fusion/flicker_synthetic/flicker_1' ] #baseDir = '/run/user/1000/gvfs/smb-share:server=192.168.0.253,share=data/Master/datasets/' #inputDirs = [ #args.inputDir #os.path.join(baseDir, 'rgbd_dataset_freiburg1_desk/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg1_desk2/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg1_plant/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg1_room/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg1_rpy/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg1_xyz/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg2_desk/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg2_xyz/rgb/'), #os.path.join(baseDir, 'rgbd_dataset_freiburg3_long_office_household/rgb/'), #] # baseDir = '/run/user/1000/gvfs/smb-share:server=192.168.0.253,share=data/Master/datasets/test_dataset' # inputDirs = [ # os.path.join(baseDir, 'hdr_fusion', 'flicker_synthetic', 'flicker_1'), # os.path.join(baseDir, 'hdr_fusion', 'smooth_synthetic', 'flicker_2'), # os.path.join(baseDir, 'nyu_depth_v2', 'basements', 'basement_001c'), # os.path.join(baseDir, 'nyu_depth_v2', 'cafe', 'cafe_0001c'), # os.path.join(baseDir, 'nyu_depth_v2', 'classrooms', 'classroom_0014'), # os.path.join(baseDir, 'tum', 'rgbd_dataset_freiburg1_desk', 'rgb'), # os.path.join(baseDir, 'tum', 'rgbd_dataset_freiburg1_xyz', 'rgb'), # os.path.join(baseDir, 'tum', 'rgbd_dataset_freiburg2_xyz', 'rgb'), # ] for inputDir in inputDirs: args.inputDir = inputDir args.cuda = True forwardAll(model, args)
def main(): args = parse_args() logger = log.get_logger(args.log) args.logger = logger logger.info('*' * 80) logger.info('the args are the below') logger.info('*' * 80) for x in args.__dict__: logger.info(x + ',' + str(args.__dict__[x])) logger.info(cfg.config[args.dataset]) logger.info('*' * 80) os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu if not os.path.exists(args.param_dir): os.mkdir(args.param_dir) torch.manual_seed(long(time.time())) model = bdcn.BDCN(pretrain=args.pretrain, logger=logger) pytorch_total_params = sum(p.numel() for p in model.parameters() if p.requires_grad) print('TOTAL TRAINABLE PARAMS: ' + str(pytorch_total_params)) if args.complete_pretrain: model.load_state_dict(torch.load(args.complete_pretrain)) logger.info(model) train(model, args)
def main(): args = parse_args() logger = log.get_logger(args.log) args.logger = logger print('*' * 80) print('the args are the below') logger.info('*' * 80) for x in args.__dict__: print(x + ',' + str(args.__dict__[x])) print(cfg.config[args.dataset]) print('*' * 80) # os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID" os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu if not os.path.exists(args.param_dir): os.mkdir(args.param_dir) torch.manual_seed(int(time.time())) device = torch.device('cpu' if torch.cuda.device_count() == 0 else 'cuda') model = bdcn.BDCN(pretrain=args.pretrain, logger=logger).to(device) if args.complete_pretrain: model.load_state_dict( torch.load(args.complete_pretrain, map_location=device)) logger.info(model) train(model, args, device=device)
# print(os.path.join(test_root, nm)) # data = cv2.resize(data, (data.shape[1]/2, data.shape[0]/2), interpolation=cv2.INTER_LINEAR) data = np.array(imagen, np.float32) data -= mean_bgr data = data.transpose((2, 0, 1)) data = torch.from_numpy(data).float().unsqueeze(0) data = Variable(data) out = model(data) out = [torch.sigmoid(x).cpu().data.numpy()[0, 0] for x in out] img = out[-1] img = 255 * img #print(img) print('Overall Time use: ', time.time() - start_time) #a = np.reshape(out,weigh,heigh) #print(type(img)) cv2.imshow("out", img) cv2.waitKey() if __name__ == '__main__': model = bdcn.BDCN() model.load_state_dict( torch.load("bdcn_pretrained_on_bsds500.pth", map_location=torch.device('cpu'))) model.eval() #print(model) test(model)