Added new correct cropping for training and smoothing for video.

natanielruiz

2017-08-08 18a21d4b07c581a8954b08518115fb035c712b28

 code/train_resnet_bins.py

@@ -91,9 +91,13 @@
    if not os.path.exists('output/snapshots'):
        os.makedirs('output/snapshots')

    # ResNet50 with 3 outputs.
    model = hopenet.Hopenet(torchvision.models.resnet.Bottleneck, [3, 4, 6, 3], 66)
    load_filtered_state_dict(model, model_zoo.load_url(model_urls['resnet50']))
    # ResNet101 with 3 outputs
    # model = hopenet.Hopenet(torchvision.models.resnet.Bottleneck, [3, 4, 23, 3], 66)
    # ResNet50
    # model = hopenet.Hopenet(torchvision.models.resnet.Bottleneck, [3, 4, 6, 3], 66)
    # ResNet18
    model = hopenet.Hopenet(torchvision.models.resnet.BasicBlock, [2, 2, 2, 2], 66)
    load_filtered_state_dict(model, model_zoo.load_url(model_urls['resnet18']))

    print 'Loading data.'

@@ -109,15 +113,15 @@

    model.cuda(gpu)
    criterion = nn.CrossEntropyLoss()
    # optimizer = torch.optim.Adam([{'params': get_ignored_params(model), 'lr': args.lr},
    #                               {'params': get_non_ignored_params(model), 'lr': args.lr * 10}],
    #                               lr = args.lr)
    optimizer = torch.optim.Adam([{'params': get_ignored_params(model), 'lr': args.lr},
                                  {'params': get_non_ignored_params(model), 'lr': args.lr * 10}],
                                  lr = args.lr)
    # optimizer = torch.optim.SGD([{'params': get_ignored_params(model), 'lr': args.lr},
    #                               {'params': get_non_ignored_params(model), 'lr': args.lr}],
    #                               lr = args.lr, momentum=0.9)
    optimizer = torch.optim.RMSprop([{'params': get_ignored_params(model), 'lr': args.lr},
                                  {'params': get_non_ignored_params(model), 'lr': args.lr * 10}],
                                  lr = args.lr)
    # optimizer = torch.optim.RMSprop([{'params': get_ignored_params(model), 'lr': args.lr},
    #                               {'params': get_non_ignored_params(model), 'lr': args.lr * 10}],
    #                               lr = args.lr)

    print 'Ready to train network.'

@@ -147,7 +151,7 @@
        if epoch % 1 == 0 and epoch < num_epochs - 1:
            print 'Taking snapshot...'
            torch.save(model.state_dict(),
            'output/snapshots/resnet50_binned_RMSprop_epoch_' + str(epoch+1) + '.pkl')
            'output/snapshots/resnet18_cr_epoch_'+ str(epoch+1) + '.pkl')

    # Save the final Trained Model
    torch.save(model.state_dict(), 'output/snapshots/resnet50_binned_RMSprop_epoch_' + str(epoch+1) + '.pkl')
    torch.save(model.state_dict(), 'output/snapshots/resnet18_cr_epoch_' + str(epoch+1) + '.pkl')