segment_st.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import cv2
import argparse
import json
import logging
import math
import os
import pdb
from os.path import exists, join, split
import threading

import time

import numpy as np
import shutil

import sys
from PIL import Image
import torch
from torch import nn
import torch.backends.cudnn as cudnn
import torch.optim as optim
from torchvision import datasets, transforms
from torch.autograd import Variable

import drn
import data_transforms as transforms

try:
    from modules import batchnormsync
except ImportError:
    pass

FORMAT = "[%(asctime)-15s %(filename)s:%(lineno)d %(funcName)s] %(message)s"
logging.basicConfig(format=FORMAT)
logger = logging.getLogger(__name__)
logger.setLevel(logging.DEBUG)


CITYSCAPE_PALETTE = np.asarray([
    [128, 64, 128],
    [244, 35, 232],
    [70, 70, 70],
    [102, 102, 156],
    [190, 153, 153],
    [153, 153, 153],
    [250, 170, 30],
    [220, 220, 0],
    [107, 142, 35],
    [152, 251, 152],
    [70, 130, 180],
    [220, 20, 60],
    [255, 0, 0],
    [0, 0, 142],
    [0, 0, 70],
    [0, 60, 100],
    [0, 80, 100],
    [0, 0, 230],
    [119, 11, 32],
    [0, 0, 0]], dtype=np.uint8)

def downsampling(x, size=None, scale=None, mode='nearest'):
    if size is None:
        size = (int(scale * x.size(2)) , int(scale * x.size(3)))
    h = torch.arange(0,size[0]) / (size[0] - 1) * 2 - 1
    w = torch.arange(0,size[1]) / (size[1] - 1) * 2 - 1
    grid = torch.zeros(size[0] , size[1] , 2)
    grid[: , : , 0] = w.unsqueeze(0).repeat(size[0] , 1)
    grid[: , : , 1] = h.unsqueeze(0).repeat(size[1] , 1).transpose(0 , 1)
    grid = grid.unsqueeze(0).repeat(x.size(0),1,1,1)
    if x.is_cuda:
        grid = grid.cuda()
    embed()
    return torch.nn.functional.grid_sample(x , grid , mode = mode)

def fill_up_weights(up):
    w = up.weight.data
    f = math.ceil(w.size(2) / 2)
    c = (2 * f - 1 - f % 2) / (2. * f)
    for i in range(w.size(2)):
        for j in range(w.size(3)):
            w[0, 0, i, j] = \
                (1 - math.fabs(i / f - c)) * (1 - math.fabs(j / f - c))
    for c in range(1, w.size(0)):
        w[c, 0, :, :] = w[0, 0, :, :]


class DRNSeg(nn.Module):
    def __init__(self, model_name, classes, pretrained_model=None,
                 pretrained=True, use_torch_up=False):
        super(DRNSeg, self).__init__()
        model = drn.__dict__.get(model_name)(
            pretrained=pretrained, num_classes=1000)
        pmodel = nn.DataParallel(model)
        if pretrained_model is not None:
            pmodel.load_state_dict(pretrained_model)
        self.base = nn.Sequential(*list(model.children())[:-2])

        self.dc = nn.Conv2d(512 , 512 , kernel_size = 3 , dilation = 8 , padding = 8 , bias = False)
        self.dc_bn = nn.BatchNorm2d(512) 
        self.drop = nn.Dropout2d(0.5)    
        self.classifier = nn.Conv2d(512 , classes , kernel_size = 1 , bias = False)
        self.softmax = nn.LogSoftmax()
        self.relu = nn.ReLU(inplace=True)

    def forward(self, x):
        x = self.base(x)
        x = self.relu(self.dc_bn(self.dc(x)))
        x = self.drop(x)
        x = self.classifier(x)
        return self.softmax(x), x

    def optim_base_parameters(self, memo=None):
        for param in self.base.parameters():
            yield param
    
    def optim_seg_parameters(self, memo=None):
        for param in self.dc.parameters():
            yield param
        for param in self.dc_bn.parameters():
            yield param
        for param in self.classifier.parameters():
            yield param


class SegList(torch.utils.data.Dataset):
    def __init__(self, data_dir, phase, transforms, list_dir=None,
                 out_name=False):
        self.list_dir = data_dir if list_dir is None else list_dir
        self.data_dir = data_dir
        self.out_name = out_name
        self.phase = phase
        self.transforms = transforms
        self.image_list = None
        self.label_list = None
        self.bbox_list = None
        self.read_lists()

    def __getitem__(self, index):
        data = [Image.open(join(self.data_dir, self.image_list[index]))]
        data = np.array(data[0])
        #print(data.shape)
        if len(data.shape) == 2:
            data = np.stack([data , data , data] , axis = 2)
        data = [Image.fromarray(data)]
        if self.label_list is not None:
       	    data.append(Image.open(join(self.data_dir, self.label_list[index])))
        data = list(self.transforms(*data))
        if self.out_name:
            if self.label_list is None:
                data.append(data[0][0, :, :])
            data.append(self.image_list[index])
        return tuple(data)

    def __len__(self):
        return len(self.image_list)

    def read_lists(self):
        image_path = join(self.list_dir, self.phase + '_images.txt')
        label_path = join(self.list_dir, self.phase + '_labels.txt')
        assert exists(image_path)
        self.image_list = [line.strip() for line in open(image_path, 'r')]
        if exists(label_path):
            self.label_list = [line.strip() for line in open(label_path, 'r')]
            assert len(self.image_list) == len(self.label_list)


class SegListMS(torch.utils.data.Dataset):
    def __init__(self, data_dir, phase, transforms, scales, list_dir=None):
        self.list_dir = data_dir if list_dir is None else list_dir
        self.data_dir = data_dir
        self.phase = phase
        self.transforms = transforms
        self.image_list = None
        self.label_list = None
        self.bbox_list = None
        self.read_lists()
        self.scales = scales

    def __getitem__(self, index):
        data = [Image.open(join(self.data_dir, self.image_list[index]))]
        w, h = data[0].size
        data = np.array(data[0])
        if len(data.shape) == 2:
            data = np.stack([data , data , data] , axis = 2)
        data = [Image.fromarray(data)]
        if self.label_list is not None:
            data.append(Image.open(join(self.data_dir, self.label_list[index])))
        # data = list(self.transforms(*data))
        out_data = list(self.transforms(*data))
        ms_images = [self.transforms(data[0].resize((int(w * s), int(h * s)),
                                                    Image.BICUBIC))[0]
                     for s in self.scales]
        out_data.append(self.image_list[index])
        out_data.extend(ms_images)
        return tuple(out_data)

    def __len__(self):
        return len(self.image_list)

    def read_lists(self):
        image_path = join(self.list_dir, self.phase + '_images.txt')
        label_path = join(self.list_dir, self.phase + '_labels.txt')
        assert exists(image_path)
        self.image_list = [line.strip() for line in open(image_path, 'r')]
        if exists(label_path):
            self.label_list = [line.strip() for line in open(label_path, 'r')]
            assert len(self.image_list) == len(self.label_list)


def validate(val_loader, model, criterion, eval_score=None, print_freq=10):
    batch_time = AverageMeter()
    losses = AverageMeter()
    score = AverageMeter()

    # switch to evaluate mode
    model.eval()

    end = time.time()
    for i, (input, target) in enumerate(val_loader):
        
        small_target = torch.zeros(int(target.size(0)) , int(target.size(1)/8) , int(target.size(2)/8))
        for index in range(0,target.size(0)):
            temp = target[index , : , :]
            temp = cv2.resize(temp.numpy(),(int(target.size(1)/8) , int(target.size(2)/8)), interpolation=cv2.INTER_NEAREST)
            temp = torch.Tensor(temp)
            small_target[index,:,:] = temp
        target = small_target
        target = target.long()
        
        if type(criterion) in [torch.nn.modules.loss.L1Loss,
                               torch.nn.modules.loss.MSELoss]:
            target = target.float()

        with torch.no_grad():
            input = input.cuda()
            target = target.cuda(async=True)
            input_var = torch.autograd.Variable(input, volatile=True)
            target_var = torch.autograd.Variable(target, volatile=True)

            # compute output
            output = model(input_var)[0]
            loss = criterion(output, target_var)

            # measure accuracy and record loss
            # prec1, prec5 = accuracy(output.data, target, topk=(1, 5))
            losses.update(loss.data[0], input.size(0))
            if eval_score is not None:
                score.update(eval_score(output, target_var), input.size(0))

        # measure elapsed time
        batch_time.update(time.time() - end)
        end = time.time()

        if i % print_freq == 0:
            logger.info('Test: [{0}/{1}]\t'
                        'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
                        'Loss {loss.val:.4f} ({loss.avg:.4f})\t'
                        'Score {score.val:.3f} ({score.avg:.3f})'.format(
                i, len(val_loader), batch_time=batch_time, loss=losses,
                score=score))

    logger.info(' * Score {top1.avg:.3f}'.format(top1=score))

    return score.avg


class AverageMeter(object):
    """Computes and stores the average and current value"""
    def __init__(self):
        self.reset()

    def reset(self):
        self.val = 0
        self.avg = 0
        self.sum = 0
        self.count = 0

    def update(self, val, n=1):
        self.val = val
        self.sum += val * n
        self.count += n
        self.avg = self.sum / self.count


def accuracy(output, target):
    """Computes the precision@k for the specified values of k"""
    # batch_size = target.size(0) * target.size(1) * target.size(2)
    _, pred = output.max(1)
    pred = pred.view(1, -1)
    target = target.view(1, -1)
    correct = pred.eq(target)
    correct = correct[target != 255]
    correct = correct.view(-1)
    score = correct.float().sum(0).mul(100.0 / correct.size(0))
    return score.data[0]


def train(train_loader, model, criterion, optimizer, epoch,
          eval_score=None, print_freq=1):
    batch_time = AverageMeter()
    data_time = AverageMeter()
    losses = AverageMeter()
    scores = AverageMeter()

    # switch to train mode
    model.train()

    end = time.time()

    for i, (input, target) in enumerate(train_loader):
        # measure data loading time
        data_time.update(time.time() - end)
        #embed()
        small_target = torch.zeros(int(target.size(0)) , int(target.size(1)/8) , int(target.size(2)/8))
        for index in range(0,target.size(0)):
            temp = target[index , : , :]
            temp = cv2.resize(temp.numpy(),(int(target.size(1)/8) , int(target.size(2)/8)), interpolation=cv2.INTER_NEAREST)
            temp = torch.Tensor(temp)
            small_target[index,:,:] = temp
        target = small_target
        target = target.long()

        if type(criterion) in [torch.nn.modules.loss.L1Loss,
                               torch.nn.modules.loss.MSELoss]:
            target = target.float()

        input = input.cuda()
        target = target.cuda(async=True)
        input_var = torch.autograd.Variable(input)
        target_var = torch.autograd.Variable(target)

        # compute output
        output = model(input_var)[0]
        loss = criterion(output, target_var)

        # measure accuracy and record loss
        # prec1, prec5 = accuracy(output.data, target, topk=(1, 5))
        losses.update(loss.data[0], input.size(0))
        if eval_score is not None:
            scores.update(eval_score(output, target_var), input.size(0))

        # compute gradient and do SGD step
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        # measure elapsed time
        batch_time.update(time.time() - end)
        end = time.time()

        if i % print_freq == 0:
            logger.info('Epoch: [{0}][{1}/{2}]\t'
                        'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
                        'Data {data_time.val:.3f} ({data_time.avg:.3f})\t'
                        'Loss {loss.val:.4f} ({loss.avg:.4f})\t'
                        'Score {top1.val:.3f} ({top1.avg:.3f})'.format(
                epoch, i, len(train_loader), batch_time=batch_time,
                data_time=data_time, loss=losses, top1=scores))


def save_checkpoint(state, is_best, filename='checkpoint.pth.tar'):
    torch.save(state, filename)
    if is_best:
        shutil.copyfile(filename, 'model_best.pth.tar')


def train_seg(args):
    batch_size = args.batch_size
    num_workers = args.workers
    crop_size = args.crop_size

    print(' '.join(sys.argv))

    for k, v in args.__dict__.items():
        print(k, ':', v)

    single_model = DRNSeg(args.arch, args.classes, None,
                          pretrained=True)
    print(single_model)
    if args.pretrained:
        #single_model.load_state_dict(torch.load(args.pretrained))
        checkpoint = torch.load(args.pretrained)
        #del checkpoint['state_dict']['module.fc.weight']
        #del checkpoint['state_dict']['module.fc.bias']
        for name, param in checkpoint['state_dict'].items():
            name = name[7:]
            single_model.state_dict()[name].copy_(param)
    model = torch.nn.DataParallel(single_model).cuda()
    criterion = nn.NLLLoss2d(ignore_index=255)

    criterion.cuda()

    # Data loading code
    data_dir = args.data_dir
    info = json.load(open(join(data_dir, 'info.json'), 'r'))
    normalize = transforms.Normalize(mean=info['mean'],
                                     std=info['std'])
    t = []
    if args.random_rotate > 0:
        t.append(transforms.RandomRotate(args.random_rotate))
    if args.random_scale > 0:
        t.append(transforms.RandomScale(args.random_scale))
    t.extend([transforms.RandomCrop(crop_size),
              transforms.RandomHorizontalFlip(),
              transforms.ToTensor(),
              normalize])
    train_loader = torch.utils.data.DataLoader(
        SegList(data_dir, 'train', transforms.Compose(t)),
        batch_size=batch_size, shuffle=True, num_workers=num_workers,
        pin_memory=True, drop_last=True
    )
    val_loader = torch.utils.data.DataLoader(
        SegList(data_dir, 'val', transforms.Compose([
            transforms.RandomCrop(crop_size),
            transforms.ToTensor(),
            normalize,
        ])),
        batch_size=1, shuffle=False, num_workers=num_workers,
        pin_memory=True, drop_last=True
    )

    # define loss function (criterion) and pptimizer
    #optimizer = torch.optim.SGD(single_model.optim_parameters(),
    #                            args.lr,
    #                            momentum=args.momentum,
    #                            weight_decay=args.weight_decay)
    optimizer = torch.optim.SGD([{'params':single_model.optim_base_parameters()},{'params':single_model.optim_seg_parameters(),'lr':args.lr*10}],
                                args.lr,
                                momentum=args.momentum,
                                weight_decay=args.weight_decay)

    cudnn.benchmark = True
    best_prec1 = 0
    start_epoch = 0

    # optionally resume from a checkpoint
    if args.resume:
        if os.path.isfile(args.resume):
            print("=> loading checkpoint '{}'".format(args.resume))
            checkpoint = torch.load(args.resume)
            start_epoch = checkpoint['epoch']
            best_prec1 = checkpoint['best_prec1']
            model.load_state_dict(checkpoint['state_dict'])
            print("=> loaded checkpoint '{}' (epoch {})"
                  .format(args.resume, checkpoint['epoch']))
        else:
            print("=> no checkpoint found at '{}'".format(args.resume))

    if args.evaluate:
        validate(val_loader, model, criterion, eval_score=accuracy)
        return

    for epoch in range(start_epoch, args.epochs):
        lr = adjust_learning_rate(args, optimizer, epoch)
        logger.info('Epoch: [{0}]\tlr {1:.06f}'.format(epoch, lr))
        # train for one epoch
        train(train_loader, model, criterion, optimizer, epoch,
              eval_score=accuracy)

        # evaluate on validation set
        prec1 = validate(val_loader, model, criterion, eval_score=accuracy)

        is_best = prec1 > best_prec1
        best_prec1 = max(prec1, best_prec1)
        checkpoint_path = 'checkpoint_latest.pth.tar'
        save_checkpoint({
            'epoch': epoch + 1,
            'arch': args.arch,
            'state_dict': model.state_dict(),
            'best_prec1': best_prec1,
        }, is_best, filename=checkpoint_path)
        if (epoch + 1) % 10 == 0:
            history_path = 'checkpoint_{:03d}.pth.tar'.format(epoch + 1)
            shutil.copyfile(checkpoint_path, history_path)


def adjust_learning_rate(args, optimizer, epoch):
    """Sets the learning rate to the initial LR decayed by 10 every 30 epochs"""
    if args.lr_mode == 'step':
        lr = args.lr * (0.1 ** (epoch // args.step))
    elif args.lr_mode == 'poly':
        lr = args.lr * (1 - epoch / args.epochs) ** 0.9
    else:
        raise ValueError('Unknown lr mode {}'.format(args.lr_mode))

    for param_group in optimizer.param_groups:
        param_group['lr'] = lr
    return lr


def fast_hist(pred, label, n):
    k = (label >= 0) & (label < n)
    return np.bincount(
        n * label[k].astype(int) + pred[k], minlength=n ** 2).reshape(n, n)


def per_class_iu(hist):
    return np.diag(hist) / (hist.sum(1) + hist.sum(0) - np.diag(hist))


def save_output_images(predictions, filenames, output_dir):
    """
    Saves a given (B x C x H x W) into an image file.
    If given a mini-batch tensor, will save the tensor as a grid of images.
    """
    # pdb.set_trace()
    for ind in range(len(filenames)):
        im = Image.fromarray(predictions[ind].astype(np.uint8))
        fn = os.path.join(output_dir, filenames[ind][:-4] + '.png')
        out_dir = split(fn)[0]
        if not exists(out_dir):
            os.makedirs(out_dir)
        im.save(fn)


def save_colorful_images(predictions, filenames, output_dir, palettes):
   """
   Saves a given (B x C x H x W) into an image file.
   If given a mini-batch tensor, will save the tensor as a grid of images.
   """
   for ind in range(len(filenames)):
       #im = Image.fromarray(palettes[predictions[ind].squeeze()])
       fn = os.path.join(output_dir, filenames[ind][:-4] + '.png')
       out_dir = split(fn)[0]
       if not exists(out_dir):
           os.makedirs(out_dir)
       im.save(fn)


def test(eval_data_loader, model, num_classes,
         output_dir='pred', has_gt=True, save_vis=False):
    model.eval()
    batch_time = AverageMeter()
    data_time = AverageMeter()
    end = time.time()
    hist = np.zeros((num_classes, num_classes))
    for iter, (image, label, name) in enumerate(eval_data_loader):
        data_time.update(time.time() - end)
        with torch.no_grad():
            image_var = Variable(image, requires_grad=False, volatile=True)
            final = model(image_var)[0]
            _, pred = torch.max(final, 1)
            pred = pred.cpu().data.numpy()
        batch_time.update(time.time() - end)
        if save_vis:
            save_output_images(pred, name, output_dir)
            #save_colorful_images(pred, name, output_dir + '_color',
            #                     CITYSCAPE_PALETTE)
        if has_gt:
            label = label.numpy()
            hist += fast_hist(pred.flatten(), label.flatten(), num_classes)
            logger.info('===> mAP {mAP:.3f}'.format(
                mAP=round(np.nanmean(per_class_iu(hist)) * 100, 2)))
        end = time.time()
        logger.info('Eval: [{0}/{1}]\t'
                    'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
                    'Data {data_time.val:.3f} ({data_time.avg:.3f})\t'
                    .format(iter, len(eval_data_loader), batch_time=batch_time,
                            data_time=data_time))
    if has_gt: #val
        ious = per_class_iu(hist) * 100
        logger.info(' '.join('{:.03f}'.format(i) for i in ious))
        return round(np.nanmean(ious), 2)


def resize_4d_tensor(tensor, width, height):
    tensor_cpu = tensor.cpu().numpy()
    if tensor.size(2) == height and tensor.size(3) == width:
        return tensor_cpu
    out_size = (tensor.size(0), tensor.size(1), height, width)
    out = np.empty(out_size, dtype=np.float32)

    def resize_one(i, j):
        out[i, j] = np.array(
            Image.fromarray(tensor_cpu[i, j]).resize(
                (width, height), Image.BILINEAR))

    def resize_channel(j):
        for i in range(tensor.size(0)):
            out[i, j] = np.array(
                Image.fromarray(tensor_cpu[i, j]).resize(
                    (width, height), Image.BILINEAR))

    # workers = [threading.Thread(target=resize_one, args=(i, j))
    #            for i in range(tensor.size(0)) for j in range(tensor.size(1))]

    workers = [threading.Thread(target=resize_channel, args=(j,))
               for j in range(tensor.size(1))]
    for w in workers:
        w.start()
    for w in workers:
        w.join()
    # for i in range(tensor.size(0)):
    #     for j in range(tensor.size(1)):
    #         out[i, j] = np.array(
    #             Image.fromarray(tensor_cpu[i, j]).resize(
    #                 (w, h), Image.BILINEAR))
    # out = tensor.new().resize_(*out.shape).copy_(torch.from_numpy(out))
    return out


def test_ms(eval_data_loader, model, num_classes, scales,
            output_dir='pred', has_gt=True, save_vis=False):
    model.eval()
    batch_time = AverageMeter()
    data_time = AverageMeter()
    end = time.time()
    hist = np.zeros((num_classes, num_classes))
    num_scales = len(scales)
    for iter, input_data in enumerate(eval_data_loader):
        data_time.update(time.time() - end)
        if has_gt:
            name = input_data[2]
            label = input_data[1]
        else:
            name = input_data[1]
        h, w = input_data[0].size()[2:4]
        images = [input_data[0]]
        images.extend(input_data[-num_scales:])
        # pdb.set_trace()
        outputs = []
        with torch.no_grad():
            for image in images:
                image_var = Variable(image, requires_grad=False)
                final = model(image_var)[0]
                outputs.append(final.data)
            final = sum([resize_4d_tensor(out, w, h) for out in outputs])
            # _, pred = torch.max(torch.from_numpy(final), 1)
            # pred = pred.cpu().numpy()
            pred = final.argmax(axis=1)
        batch_time.update(time.time() - end)
        if save_vis:
            save_output_images(pred, name, output_dir)
            #save_colorful_images(pred, name, output_dir + '_color',
            #                     CITYSCAPE_PALETTE)
        if has_gt:
            label = label.numpy()
            hist += fast_hist(pred.flatten(), label.flatten(), num_classes)
            logger.info('===> mAP {mAP:.3f}'.format(
                mAP=round(np.nanmean(per_class_iu(hist)) * 100, 2)))
        end = time.time()
        logger.info('Eval: [{0}/{1}]\t'
                    'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
                    'Data {data_time.val:.3f} ({data_time.avg:.3f})\t'
                    .format(iter, len(eval_data_loader), batch_time=batch_time,
                            data_time=data_time))
    if has_gt: #val
        ious = per_class_iu(hist) * 100
        logger.info(' '.join('{:.03f}'.format(i) for i in ious))
        return round(np.nanmean(ious), 2)


def test_seg(args):
    batch_size = args.batch_size
    num_workers = args.workers
    phase = args.phase

    for k, v in args.__dict__.items():
        print(k, ':', v)

    single_model = DRNSeg(args.arch, args.classes, pretrained_model=None,
                          pretrained=False)
    if args.pretrained:
        single_model.load_state_dict(torch.load(args.pretrained))
    model = torch.nn.DataParallel(single_model).cuda()

    data_dir = args.data_dir
    info = json.load(open(join(data_dir, 'info.json'), 'r'))
    normalize = transforms.Normalize(mean=info['mean'], std=info['std'])
    scales = [0.5, 0.75, 1.25, 1.5, 1.75]
    #scales = [1]
    if args.ms:
        dataset = SegListMS(data_dir, phase, transforms.Compose([
            transforms.ToTensor(),
            normalize,
        ]), scales)
    else:
        dataset = SegList(data_dir, phase, transforms.Compose([
            transforms.ToTensor(),
            normalize,
        ]), out_name=True)
    test_loader = torch.utils.data.DataLoader(
        dataset,
        batch_size=batch_size, shuffle=False, num_workers=num_workers,
        pin_memory=False
    )

    cudnn.benchmark = True

    # optionally resume from a checkpoint
    start_epoch = 0
    if args.resume:
        if os.path.isfile(args.resume):
            logger.info("=> loading checkpoint '{}'".format(args.resume))
            checkpoint = torch.load(args.resume)
            start_epoch = checkpoint['epoch']
            best_prec1 = checkpoint['best_prec1']
            model.load_state_dict(checkpoint['state_dict'])
            logger.info("=> loaded checkpoint '{}' (epoch {})"
                  .format(args.resume, checkpoint['epoch']))
        else:
            logger.info("=> no checkpoint found at '{}'".format(args.resume))

    out_dir = '{}_{:03d}_{}'.format(args.arch, start_epoch, phase)
    if len(args.test_suffix) > 0:
        out_dir += '_' + args.test_suffix
    if args.ms:
        out_dir += '_ms'

    if args.ms:
        mAP = test_ms(test_loader, model, args.classes, save_vis=True,
                      has_gt=phase != 'test' or args.with_gt,
                      output_dir=out_dir,
                      scales=scales)
    else:
        mAP = test(test_loader, model, args.classes, save_vis=True,
                   has_gt=phase != 'test' or args.with_gt, output_dir=out_dir)
    logger.info('mAP: %f', mAP)


def parse_args():
    # Training settings
    parser = argparse.ArgumentParser(description='')
    parser.add_argument('cmd', choices=['train', 'test'])
    parser.add_argument('-d', '--data-dir', default=None)
    parser.add_argument('-c', '--classes', default=0, type=int)
    parser.add_argument('-s', '--crop-size', default=0, type=int)
    parser.add_argument('--step', type=int, default=200)
    parser.add_argument('--arch')
    parser.add_argument('--batch-size', type=int, default=64, metavar='N',
                        help='input batch size for training (default: 64)')
    parser.add_argument('--epochs', type=int, default=10, metavar='N',
                        help='number of epochs to train (default: 10)')
    parser.add_argument('--lr', type=float, default=0.01, metavar='LR',
                        help='learning rate (default: 0.01)')
    parser.add_argument('--lr-mode', type=str, default='step')
    parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
                        help='SGD momentum (default: 0.9)')
    parser.add_argument('--weight-decay', '--wd', default=1e-4, type=float,
                        metavar='W', help='weight decay (default: 1e-4)')
    parser.add_argument('-e', '--evaluate', dest='evaluate',
                        action='store_true',
                        help='evaluate model on validation set')
    parser.add_argument('--resume', default='', type=str, metavar='PATH',
                        help='path to latest checkpoint (default: none)')
    parser.add_argument('--pretrained', dest='pretrained',
                        default='', type=str, metavar='PATH',
                        help='use pre-trained model')
    parser.add_argument('-j', '--workers', type=int, default=8)
    parser.add_argument('--load-release', dest='load_rel', default=None)
    parser.add_argument('--phase', default='val')
    parser.add_argument('--random-scale', default=0, type=float)
    parser.add_argument('--random-rotate', default=0, type=int)
    parser.add_argument('--bn-sync', action='store_true')
    parser.add_argument('--ms', action='store_true',
                        help='Turn on multi-scale testing')
    parser.add_argument('--with-gt', action='store_true')
    parser.add_argument('--test-suffix', default='', type=str)
    args = parser.parse_args()

    assert args.data_dir is not None
    assert args.classes > 0

    print(' '.join(sys.argv))
    print(args)

    if args.bn_sync:
        drn.BatchNorm = batchnormsync.BatchNormSync

    return args


def main():
    args = parse_args()
    if args.cmd == 'train':
        train_seg(args)
    elif args.cmd == 'test':
        test_seg(args)


if __name__ == '__main__':
    main()