def main(): """ visualizes the dataset """ from common.vis_utils import show_batch, show_stereo_batch from torchvision.utils import make_grid import torchvision.transforms as transforms seq = 'chess' mode = 2 num_workers = 6 transform = transforms.Compose([ transforms.Scale(256), transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) ]) dset = SevenScenes(seq, '../data/deepslam_data/7Scenes', True, transform, mode=mode) print( ('Loaded 7Scenes sequence {:s}, length = {:d}'.format(seq, len(dset)))) data_loader = data.DataLoader(dset, batch_size=10, shuffle=True, num_workers=num_workers) batch_count = 0 N = 2 for batch in data_loader: print(('Minibatch {:d}'.format(batch_count))) if mode < 2: show_batch(make_grid(batch[0], nrow=1, padding=25, normalize=True)) elif mode == 2: lb = make_grid(batch[0][0], nrow=1, padding=25, normalize=True) rb = make_grid(batch[0][1], nrow=1, padding=25, normalize=True) show_stereo_batch(lb, rb) batch_count += 1 if batch_count >= N: break
def main(): """ visualizes the dataset """ from common.vis_utils import show_batch, show_stereo_batch from torchvision.utils import make_grid import torchvision.transforms as transforms mode = int(sys.argv[1]) num_workers = 6 transform = transforms.Compose([ transforms.Scale(256), transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) ]) concatenate_inputs = (mode == 3) if mode == 0: input_types = 'left' elif mode == 1 and not dual_output: input_types = 'depth' elif mode == 2 and not dual_output: input_types = ['left', 'depth'] elif mode == 3: input_types = ['left', 'label_colorized'] dset = DeepLoc('../data/deepslam_data/DeepLoc', True, transform, input_types=input_types, output_types=[], concatenate_inputs=concatenate_inputs) print('Loaded DeepLoc, length = {:d}'.format(len(dset))) data_loader = data.DataLoader(dset, batch_size=10, shuffle=True, num_workers=num_workers) batch_count = 0 N = 2 for batch in data_loader: print('Minibatch {:d}'.format(batch_count)) if mode < 2: show_batch(make_grid(batch[0], nrow=1, padding=25, normalize=True)) elif mode == 2: lb = make_grid(batch[0][0], nrow=1, padding=25, normalize=True) rb = make_grid(batch[0][1], nrow=1, padding=25, normalize=True) show_stereo_batch(lb, rb) elif mode == 3: print(len(batch)) for elm in batch: print(elm.shape) lb = make_grid(batch[0][0][:3], nrow=1, padding=25, normalize=True) rb = make_grid(batch[0][0][3:], nrow=1, padding=25, normalize=True) show_stereo_batch(lb, rb) batch_count += 1 if batch_count >= N: break