trainval_fcn.py

# --------------------------------------------------------
# Pytorch multi-GPU Faster R-CNN
# Licensed under The MIT License [see LICENSE for details]
# Written by Jiasen Lu, Jianwei Yang, based on code from Ross Girshick
# --------------------------------------------------------
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import _init_paths
import os
import sys
import numpy as np
import argparse
import pprint
import pdb
import time

import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim as optim

import torchvision
import torchvision.transforms as transforms
from torch.utils.data.sampler import Sampler

from roi_data_layer.roidb import combined_roidb
from roi_data_layer.roibatchLoader import roibatchLoader
from model.utils.config import cfg, cfg_from_file, cfg_from_list, get_output_dir
from model.utils.net_utils import weights_normal_init, save_net, load_net, \
      adjust_learning_rate, save_checkpoint, clip_gradient

from model.faster_rcnn.vgg16 import vgg16
from model.faster_rcnn.resnet import resnet
from model.faster_rcnn.resnet_fcn import resnet_fcn


def parse_args():
  """
  Parse input arguments
  """
  parser = argparse.ArgumentParser(description='Train a CADOR network')
  parser.add_argument('--dataset', dest='dataset',
                      help='training dataset',
                      default='dota', type=str)
  parser.add_argument('--net', dest='net',
                    help='vgg16, res101',
                    default='res101', type=str)  # 'vgg16'
  parser.add_argument('--start_epoch', dest='start_epoch',
                      help='starting epoch',
                      default=1, type=int)
  parser.add_argument('--epochs', dest='max_epochs',
                      help='number of epochs to train',
                      default=10, type=int)
  parser.add_argument('--disp_interval', dest='disp_interval',
                      help='number of iterations to display',
                      default=100, type=int)
  parser.add_argument('--checkpoint_interval', dest='checkpoint_interval',
                      help='number of iterations to display',
                      default=10000, type=int)

  parser.add_argument('--save_dir', dest='save_dir',
                      help='directory to save models', default="models",
                      type=str)
  parser.add_argument('--nw', dest='num_workers',
                      help='number of worker to load data',
                      default=0, type=int)
  parser.add_argument('--cuda', dest='cuda',
                      default=True,  #SL 使用GPU
                      help='whether use CUDA',
                      action='store_true')
  parser.add_argument('--ls', dest='large_scale',
                      help='whether use large imag scale',
                      action='store_true')                      
  parser.add_argument('--mGPUs', dest='mGPUs',
                      help='whether use multiple GPUs',
                      action='store_true')
  parser.add_argument('--bs', dest='batch_size',
                      help='batch_size',
                      default=4, type=int) #32
  parser.add_argument('--cag', dest='class_agnostic',
                      help='whether perform class_agnostic bbox regression',
                      action='store_true')

# config optimization
  parser.add_argument('--o', dest='optimizer',
                      help='training optimizer',
                      default="sgd", type=str)
  parser.add_argument('--lr', dest='lr',
                      help='starting learning rate',
                      default=4e-3, type=float)  # 0.001
  parser.add_argument('--lr_decay_step', dest='lr_decay_step',
                      help='step to do learning rate decay, unit is epoch',
                      default=8, type=int)
  parser.add_argument('--lr_decay_gamma', dest='lr_decay_gamma',
                      help='learning rate decay ratio',
                      default=0.1, type=float) #

# set training session
  parser.add_argument('--s', dest='session',
                      help='training session',
                      default=1, type=int)

# resume trained model
  parser.add_argument('--r', dest='resume',
                      help='resume checkpoint or not',
                      default=False, type=bool)
  parser.add_argument('--checksession', dest='checksession',
                      help='checksession to load model',
                      default=1, type=int)
  parser.add_argument('--checkepoch', dest='checkepoch',
                      help='checkepoch to load model',
                      default=1, type=int)
  parser.add_argument('--checkpoint', dest='checkpoint',
                      help='checkpoint to load model',
                      default=0, type=int)
# log and diaplay
  parser.add_argument('--use_tfb', dest='use_tfboard',
                      help='whether use tensorboard',
                      default=True,
                      action='store_true')

  args = parser.parse_args()
  return args


class sampler(Sampler):
  def __init__(self, train_size, batch_size):
    self.num_data = train_size
    self.num_per_batch = int(train_size / batch_size)
    self.batch_size = batch_size
    self.range = torch.arange(0,batch_size).view(1, batch_size).long()
    self.leftover_flag = False
    if train_size % batch_size:
      self.leftover = torch.arange(self.num_per_batch*batch_size, train_size).long()
      self.leftover_flag = True

  def __iter__(self):
    rand_num = torch.randperm(self.num_per_batch).view(-1,1) * self.batch_size
    self.rand_num = rand_num.expand(self.num_per_batch, self.batch_size) + self.range

    self.rand_num_view = self.rand_num.view(-1)

    if self.leftover_flag:
      self.rand_num_view = torch.cat((self.rand_num_view, self.leftover),0)

    return iter(self.rand_num_view)

  def __len__(self):
    return self.num_data


###########################################
#
#
###########################################

class SquaredErrorLoss(torch.nn.Module):
    def __init__(self):
        super(SquaredErrorLoss, self).__init__()
        
    def forward(self, y_fcn, im_data, im_info, gt_boxes, num_boxes): # X, Y, Classes
    
        n = im_data.shape[1]
        index = torch.tensor((range(0,n))).to(im_data.device)
        
        Error_sum = torch.tensor(0.0)
        #pdb.set_trace()
        
        box = [0,0,600,600]
        Error_sum = (im_data.repeat(1,16,1,1) - y_fcn).pow(2).sum() \
                      /im_data.shape[0]/ (n * (box[3] - box[1]) * (box[2] - box[0]))
        
        num = 0
        for batch in range(len(num_boxes)):
            for gt_box in gt_boxes[batch, 0:num_boxes[batch], :].type(torch.int64):
                box, clg = gt_box[0:4], gt_box[4]
                
                # gt_box [x1,y1,x2,y2,clg]
                # print(batch, box, clg, (n * (box[3] - box[1]) * (box[2] - box[0])))
                Y_class = y_fcn[batch, index+n*clg, :,:]
                
                # box = [0,0,600,600] #SL
                #pdb.set_trace()
                
                # plt.imshow(1+(im_data[batch,:,box[1]:box[3], box[0]:box[2]])
                #            .permute([1,2,0]).cpu().detach().numpy()/100)
                
                Error_value = torch.pow(im_data[batch,:,box[1]:box[3], box[0]:box[2]] - 
                                  Y_class[:,box[1]:box[3], box[0]:box[2]], 2).sum() \
                                  / (n * (box[3] - box[1]) * (box[2] - box[0]))
                #print(Error_value)
                Error_sum = Error_sum + Error_value
                num = num + 1
                
        Error_sum = Error_sum/ num_boxes.sum()
        
        torch.cuda.empty_cache()
        return Error_sum




if __name__ == '__main__':

  args = parse_args()

  print('Called with args:')
  print(args)

  if args.dataset == "dota": # 改为dota
      args.imdb_name = "voc_dota_train" # trainval
      args.imdbval_name = "voc_dota_test"
      args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '215']
  elif args.dataset == "pascal_voc":
      args.imdb_name = "voc_2007_trainval"
      args.imdbval_name = "voc_2007_test"
      args.set_cfgs = ['ANCHOR_SCALES', '[8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '20']
  elif args.dataset == "pascal_voc_0712":
      args.imdb_name = "voc_2007_trainval+voc_2012_trainval"
      args.imdbval_name = "voc_2007_test"
      args.set_cfgs = ['ANCHOR_SCALES', '[8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '20']
  elif args.dataset == "pascal_voc_2012":
      args.imdb_name = "voc_2012_trainval"
      args.imdbval_name = "voc_2012_test"
      args.set_cfgs = ['ANCHOR_SCALES', '[8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '20']
  elif args.dataset == "coco":
      args.imdb_name = "coco_2014_train+coco_2014_valminusminival"
      args.imdbval_name = "coco_2014_minival"
      args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '50']
  elif args.dataset == "imagenet":
      args.imdb_name = "imagenet_train"
      args.imdbval_name = "imagenet_val"
      args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '30']
  elif args.dataset == "vg":
      # train sizes: train, smalltrain, minitrain
      # train scale: ['150-50-20', '150-50-50', '500-150-80', '750-250-150', '1750-700-450', '1600-400-20']
      args.imdb_name = "vg_150-50-50_minitrain"
      args.imdbval_name = "vg_150-50-50_minival"
      args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '50']

  args.cfg_file = "cfgs/{}_ls.yml".format(args.net) if args.large_scale else "cfgs/{}.yml".format(args.net)

  if args.cfg_file is not None:
    cfg_from_file(args.cfg_file)
  if args.set_cfgs is not None:
    cfg_from_list(args.set_cfgs)

  print('Using config:')
  pprint.pprint(cfg)
  np.random.seed(cfg.RNG_SEED)

  #torch.backends.cudnn.benchmark = True
  if torch.cuda.is_available() and not args.cuda:
    print("WARNING: You have a CUDA device, so you should probably run with --cuda")

  # train set
  # -- Note: Use validation set and disable the flipped to enable faster loading.
  cfg.TRAIN.USE_FLIPPED = False #True
  cfg.USE_GPU_NMS = args.cuda
  imdb, roidb, ratio_list, ratio_index = combined_roidb(args.imdb_name)
  args.num_classes = imdb.num_classes
  train_size = len(roidb)

  print('{:d} roidb entries'.format(len(roidb)))

  output_dir = args.save_dir + "/" + args.net + "/" + args.dataset
  if not os.path.exists(output_dir):
    os.makedirs(output_dir)

  sampler_batch = sampler(train_size, args.batch_size)

  dataset = roibatchLoader(roidb, ratio_list, ratio_index, args.batch_size, \
                           imdb.num_classes, training=True)

  dataloader = torch.utils.data.DataLoader(dataset, batch_size=args.batch_size,
                            sampler=sampler_batch, num_workers=args.num_workers)
      
  # initilize the tensor holder here.
  im_data = torch.FloatTensor(1)
  im_info = torch.FloatTensor(1)
  num_boxes = torch.LongTensor(1)
  gt_boxes = torch.FloatTensor(1)
  
  # ship to cuda
  if args.cuda:
    im_data = im_data.cuda()
    im_info = im_info.cuda()
    num_boxes = num_boxes.cuda()
    gt_boxes = gt_boxes.cuda()

  # make variable
  im_data = Variable(im_data)
  im_info = Variable(im_info)
  num_boxes = Variable(num_boxes)
  gt_boxes = Variable(gt_boxes)

  if args.cuda:
    cfg.CUDA = True

  # initilize the network here.
  if args.net == 'vgg16':
    fasterRCNN = vgg16(imdb.classes, pretrained=True, class_agnostic=args.class_agnostic)
  elif args.net == 'res101':
    #fasterRCNN = resnet(imdb.classes, 101, pretrained=True, class_agnostic=args.class_agnostic)
    fasterRCNN_FCN = resnet_fcn(imdb.classes, 101, pretrained=True, class_agnostic=args.class_agnostic)
  elif args.net == 'res50':
    fasterRCNN = resnet(imdb.classes, 50, pretrained=True, class_agnostic=args.class_agnostic)
  elif args.net == 'res152':
    fasterRCNN = resnet(imdb.classes, 152, pretrained=True, class_agnostic=args.class_agnostic)
  else:
    print("network is not defined")
    pdb.set_trace()
    
  
  #fasterRCNN.create_architecture()
  fasterRCNN_FCN.create_architecture()

  # 增加 error map, 更改fasterRCNN
  num_classes = 3 * imdb.num_classes # background+15类
  model_fcn = torchvision.models.segmentation.deeplabv3_resnet50(pretrained=False, 
                                                     progress=True, 
                                                     num_classes=3 * imdb.num_classes,
                                                     aux_loss=None)
  model_fcn.classifier[4] = torch.nn.Conv2d(512, 3 * imdb.num_classes, kernel_size=1,stride=1)
  #model_fcn.load_state_dict(torch.load('DOAT_FCN-ResNet101_model_last.pth'))
  
  # fasterRCNN.RCNN_base[0] = torch.nn.Conv2d(3 * imdb.num_classes, 64, 
  #                                           kernel_size=7,stride=2,
  #                                           padding=3,bias=False)
  
  
  # FCN 损失和优化函数
  loss_fcn = SquaredErrorLoss()

  lr = cfg.TRAIN.LEARNING_RATE
  lr = args.lr
  #tr_momentum = cfg.TRAIN.MOMENTUM
  #tr_momentum = args.momentum

  params = [] # 
  for key, value in {**dict(fasterRCNN_FCN.named_parameters()),**dict(model_fcn.named_parameters())}.items():
    if value.requires_grad:
      if 'bias' in key:
        params += [{'params':[value],'lr':lr*(cfg.TRAIN.DOUBLE_BIAS + 1), \
                'weight_decay': cfg.TRAIN.BIAS_DECAY and cfg.TRAIN.WEIGHT_DECAY or 0}]
      else:
        params += [{'params':[value],'lr':lr, 'weight_decay': cfg.TRAIN.WEIGHT_DECAY}]

  if args.cuda:
    fasterRCNN_FCN.cuda()
    model_fcn.cuda()
      
  if args.optimizer == "adam":
    lr = lr * 0.1
    optimizer = torch.optim.Adam(params)

  elif args.optimizer == "sgd":
    optimizer = torch.optim.SGD(params, momentum=cfg.TRAIN.MOMENTUM)
    
  if args.resume:
    load_name = os.path.join(output_dir,
      'faster_rcnn_{}_{}_{}.pth'.format(args.checksession, args.checkepoch, args.checkpoint))
    print("loading checkpoint %s" % (load_name))
    checkpoint = torch.load(load_name)
    args.session = checkpoint['session']
    args.start_epoch = checkpoint['epoch']
    fasterRCNN.load_state_dict(checkpoint['model'])
    optimizer.load_state_dict(checkpoint['optimizer'])
    lr = optimizer.param_groups[0]['lr']
    if 'pooling_mode' in checkpoint.keys():
      cfg.POOLING_MODE = checkpoint['pooling_mode']
    print("loaded checkpoint %s" % (load_name))

  if args.mGPUs:
    fasterRCNN_FCN = nn.DataParallel(fasterRCNN_FCN)

  iters_per_epoch = int(train_size / args.batch_size)

  if args.use_tfboard:
    from tensorboardX import SummaryWriter
    logger = SummaryWriter("logs")
    
  with open('log.txt','a')as file:
      file.write("\n\n===Begin Trainning %s ===\n"%(time.strftime("%Y-%m-%d %H:%M:%S")))

  for epoch in range(args.start_epoch, args.max_epochs + 1):
    # setting to train mode
    fasterRCNN_FCN.train()
    model_fcn.train()
    loss_temp = 0
    start = time.time()

    if epoch % (args.lr_decay_step + 1) == 0:
        adjust_learning_rate(optimizer, args.lr_decay_gamma)
        lr *= args.lr_decay_gamma

    data_iter = iter(dataloader)
    for step in range(iters_per_epoch):
      data = next(data_iter)
      with torch.no_grad():
              im_data.resize_(data[0].size()).copy_(data[0])
              im_info.resize_(data[1].size()).copy_(data[1])
              gt_boxes.resize_(data[2].size()).copy_(data[2])
              num_boxes.resize_(data[3].size()).copy_(data[3])
      
      y_fcn = model_fcn(im_data)['out']
      error_map = torch.pow(torch.cat([im_data]* imdb.num_classes, axis=1) - y_fcn, 2)
      
      fcn_loss = loss_fcn(y_fcn, im_data, im_info, gt_boxes, num_boxes)/10000
      
      # fasterRCNN.zero_grad()
      # rois, cls_prob, bbox_pred, \
      # rpn_loss_cls, rpn_loss_box, \
      # RCNN_loss_cls, RCNN_loss_bbox, \
      # rois_label = fasterRCNN(im_data, im_info, gt_boxes, num_boxes) #error_map
      
      fasterRCNN_FCN.zero_grad()
      rois, cls_prob, bbox_pred, \
      rpn_loss_cls, rpn_loss_box, \
      RCNN_loss_cls, RCNN_loss_bbox, fcn_loss, \
      rois_label = fasterRCNN_FCN(im_data, im_info, gt_boxes, num_boxes)
      
      # 损失函数
      loss = rpn_loss_cls.mean() + rpn_loss_box.mean() \
             + RCNN_loss_cls.mean() + RCNN_loss_bbox.mean() + fcn_loss
      # loss = fcn_loss
      loss_temp += loss.item()
      
      if fcn_loss.item() < 0:
          pdb.set_trace()

      # backward
      optimizer.zero_grad()
      loss.backward()
      if args.net == "vgg16":
          clip_gradient(fasterRCNN, 10.)
      optimizer.step()
      
      #torch.cuda.empty_cache()
      if step % args.disp_interval == 0:
        end = time.time()
        if step > 0:
          loss_temp /= (args.disp_interval + 1)

        if args.mGPUs:
          loss_rpn_cls = rpn_loss_cls.mean().item()
          loss_rpn_box = rpn_loss_box.mean().item()
          loss_rcnn_cls = RCNN_loss_cls.mean().item()
          loss_rcnn_box = RCNN_loss_bbox.mean().item()
          loss_fcn_item = fcn_loss.item()
          fg_cnt = torch.sum(rois_label.data.ne(0))
          bg_cnt = rois_label.data.numel() - fg_cnt
        else:
          loss_rpn_cls = rpn_loss_cls.item()
          loss_rpn_box = rpn_loss_box.item()
          loss_rcnn_cls = RCNN_loss_cls.item()
          loss_rcnn_box = RCNN_loss_bbox.item()
          loss_fcn_item = fcn_loss.item()
          fg_cnt = torch.sum(rois_label.data.ne(0))
          bg_cnt = rois_label.data.numel() - fg_cnt
          
        

        print("[session %d][epoch %2d][iter %4d/%4d] loss: %.4f, lr: %.2e" \
                                % (args.session, epoch, step, iters_per_epoch, loss_temp, lr))
        print("\t\t\tfg/bg=(%d/%d), time cost: %f" % (fg_cnt, bg_cnt, end-start))
        print("\t\t\trpn_cls:%.4f, rpn_box:%.4f, rcnn_cls:%.4f, rcnn_box:%.4f, fcn_loss:%.4f" \
                      % (loss_rpn_cls, loss_rpn_box, loss_rcnn_cls, loss_rcnn_box, loss_fcn_item))

        with open('log.txt','a')as file:
            file.write("rpn_cls: %.4f, rpn_box: %.4f, rcnn_cls: %.4f, rcnn_box: %.4f, fcn_loss: %.4f\n" \
                          % (loss_rpn_cls, loss_rpn_box, loss_rcnn_cls, loss_rcnn_box, loss_fcn_item))
        #print(fcn_loss)
        if loss_fcn_item < 0:
            pdb.set_trace()
            
        if args.use_tfboard:
          info = {
            'loss': loss_temp,
            'loss_rpn_cls': loss_rpn_cls,
            'loss_rpn_box': loss_rpn_box,
            'loss_rcnn_cls': loss_rcnn_cls,
            'loss_rcnn_box': loss_rcnn_box,
            'loss_fcn_item': loss_fcn_item
          }
          logger.add_scalars("logs_s_{}/losses".format(args.session), info, (epoch - 1) * iters_per_epoch + step)

        loss_temp = 0
        start = time.time()

    
    save_name = os.path.join(output_dir, 'faster_rcnn_{}_{}_{}.pth'.format(args.session, epoch, step))
    save_checkpoint({
      'session': args.session,
      'epoch': epoch + 1,
      'model': fasterRCNN_FCN.module.state_dict() if args.mGPUs else fasterRCNN_FCN.state_dict(),
      'optimizer': optimizer.state_dict(),
      'pooling_mode': cfg.POOLING_MODE,
      'class_agnostic': args.class_agnostic,
    }, save_name)
    print('save model: {}'.format(save_name))
    
    #save_name = os.path.join(output_dir, 'fcn-resnet_{}_{}_{}.pth'.format(args.session, epoch, step))
    #torch.save(model_fcn.state_dict(), save_name)
    
  if args.use_tfboard:
    logger.close()