def execute(self, data):
output_dict = OrderedDict()
# data.shape = [N,C,D,H,W], data is a whole vide, D=the length of the video
clip_list = self.get_clip_by_stride(
data) # the length of the length is the length of the video
scores = np.empty(shape=(len(clip_list), ), dtype=np.float32)
for index, clip in enumerate(clip_list):
first_frame = clip[:, :, 0, :, :].cuda()
second_frame = clip[:, :, 1, :, :].cuda()
generated_flow, generated_frame = self.G(first_frame)
gtFlowEstim = torch.cat([first_frame, second_frame], 1)
_, gtFlow = flow_batch_estimate(self.F,
gtFlowEstim,
self.normalize.param['val'],
output_format=self.optical_format,
optical_size=self.optical_size)
score, _, _ = amc_score(second_frame, generated_frame, gtFlow,
generated_flow, self.wf, self.wi)
score = score.tolist()
scores[index] = score
result_mask = scores.gt(self.threshold)
output_dict['result_dict'] = result_mask
return output_dict
def train(self, current_step):
# Pytorch [N, C, D, H, W]
# initialize
start = time.time()
self.set_requires_grad(self.F, False)
self.set_requires_grad(self.G, True)
self.set_requires_grad(self.D, True)
self.G.train()
self.D.train()
self.F.eval()
writer = self.kwargs['writer_dict']['writer']
global_steps = self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])]
# get the data
data, anno, meta = next(self._train_loader_iter)
self.data_time.update(time.time() - start)
# base on the D to get each frame
target = data[:, :, -1, :, :].cuda() # t+1 frame
input_data = data[:, :, :-1, :, :] # 0 ~ t frame
input_last = input_data[:, :, -1, :, :].cuda() # t frame
# squeeze the D dimension to C dimension, shape comes to [N, C, H, W]
input_data = input_data.reshape(input_data.shape[0], -1,
input_data.shape[-2],
input_data.shape[-1]).cuda()
# True Process =================Start===================
#---------update optim_G ---------
self.set_requires_grad(self.D, False)
output_pred_G = self.G(input_data)
# import ipdb; ipdb.set_trace()
predFlowEstim = torch.cat([input_last, output_pred_G], 1)
gtFlowEstim = torch.cat([input_last, target], 1)
gtFlow_vis, gtFlow = flow_batch_estimate(
self.F,
gtFlowEstim,
self.normalize.param['train'],
output_format=self.config.DATASET.optical_format,
optical_size=self.config.DATASET.optical_size)
predFlow_vis, predFlow = flow_batch_estimate(
self.F,
predFlowEstim,
self.normalize.param['train'],
output_format=self.config.DATASET.optical_format,
optical_size=self.config.DATASET.optical_size)
loss_g_adv = self.GANLoss(self.D(output_pred_G), True)
loss_op = self.OpticalflowLoss(predFlow, gtFlow)
loss_int = self.IntentsityLoss(output_pred_G, target)
loss_gd = self.GradientLoss(output_pred_G, target)
loss_g_all = self.loss_lamada[
'IntentsityLoss'] * loss_int + self.loss_lamada[
'GradientLoss'] * loss_gd + self.loss_lamada[
'OpticalflowLoss'] * loss_op + self.loss_lamada[
'GANLoss'] * loss_g_adv
self.optimizer_G.zero_grad()
loss_g_all.backward()
self.optimizer_G.step()
# record
self.loss_meter_G.update(loss_g_all.detach())
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_G_scheduler.step()
#---------update optim_D ---------------
self.set_requires_grad(self.D, True)
self.optimizer_D.zero_grad()
# G_output = self.G(input)
temp_t = self.D(target)
temp_g = self.D(output_pred_G.detach())
loss_d_1 = self.GANLoss(temp_t, True)
loss_d_2 = self.GANLoss(temp_g, False)
loss_d = (loss_d_1 + loss_d_2) * 0.5
# loss_d.sum().backward()
loss_d.backward()
self.optimizer_D.step()
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_D_scheduler.step()
self.loss_meter_D.update(loss_d.detach())
# ======================End==================
self.batch_time.update(time.time() - start)
if (current_step % self.steps.param['log'] == 0):
msg = make_info_message(current_step, self.steps.param['max'],
self.kwargs['model_type'], self.batch_time,
self.config.TRAIN.batch_size,
self.data_time,
[self.loss_meter_G, self.loss_meter_D])
self.logger.info(msg)
writer.add_scalar('Train_loss_G', self.loss_meter_G.val, global_steps)
writer.add_scalar('Train_loss_D', self.loss_meter_D.val, global_steps)
if (current_step % self.steps.param['vis'] == 0):
vis_objects = OrderedDict({
'train_target':
target.detach(),
'train_output_pred_G':
output_pred_G.detach(),
'train_gtFlow':
gtFlow_vis.detach(),
'train_predFlow':
predFlow_vis.detach()
})
tensorboard_vis_images(vis_objects, writer, global_steps,
self.normalize.param['train'])
global_steps += 1
# reset start
start = time.time()
# self.saved_model = {'G':self.G, 'D':self.D}
self.saved_model['G'] = self.G
self.saved_model['D'] = self.D
# self.saved_optimizer = {'optim_G': self.optimizer_G, 'optim_D': self.optimizer_D}
self.saved_optimizer['optimizer_G'] = self.optimizer_G
self.saved_optimizer['optimizer_D'] = self.optimizer_D
# self.saved_loss = {'loss_G':self.loss_meter_G.val, 'loss_D':self.loss_meter_D.val}
self.saved_loss['loss_G'] = self.loss_meter_G.val
self.saved_loss['loss_D'] = self.loss_meter_D.val
self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])] = global_steps
def train_pcm(self, current_step):
# Pytorch [N, C, D, H, W]
# initialize
start = time.time()
if self.kwargs['parallel']:
self.set_requires_grad(self.G.module.pcm, True)
self.set_requires_grad(self.G.module.erm, False)
else:
self.set_requires_grad(self.G.pcm, True)
self.set_requires_grad(self.G.erm, False)
writer = self.kwargs['writer_dict']['writer']
global_steps = self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])]
# get the data
data, anno, meta = next(
self._train_loader_iter) # the core for dataloader
self.data_time.update(time.time() - start)
# base on the D to get each frame
target = data[:, :, -1, :, :].cuda() # t frame
pred_last = data[:, :, -2, :, :].cuda() # t-1 frame
input_data = data[:, :, :-1, :, :].cuda() # 0 ~ t-1 frame
# input_data = data.cuda() # 0 ~ t frame
# True Process =================Start===================
#---------update optim_G ---------
self.set_requires_grad(self.D, False)
output_predframe_G, _ = self.G(input_data, target)
predFlowEstim = torch.cat([pred_last, output_predframe_G], 1).cuda()
gtFlowEstim = torch.cat([pred_last, target], 1).cuda()
gtFlow_vis, gtFlow = flow_batch_estimate(
self.F,
gtFlowEstim,
self.normalize.param['train'],
output_format=self.config.DATASET.optical_format,
optical_size=self.config.DATASET.optical_size)
predFlow_vis, predFlow = flow_batch_estimate(
self.F,
predFlowEstim,
self.normalize.param['train'],
output_format=self.config.DATASET.optical_format,
optical_size=self.config.DATASET.optical_size)
loss_g_adv = self.GANLoss(self.D(output_predframe_G), True)
loss_op = self.OpticalflowSqrtLoss(predFlow, gtFlow)
loss_int = self.IntentsityLoss(output_predframe_G, target)
loss_gd = self.GradientLoss(output_predframe_G, target)
loss_g_all = self.loss_lamada['IntentsityLoss'] * loss_int + self.loss_lamada['GradientLoss'] * loss_gd + \
self.loss_lamada['OpticalflowSqrtLoss'] * loss_op + self.loss_lamada['GANLoss'] * loss_g_adv
self.optimizer_G.zero_grad()
loss_g_all.backward()
self.optimizer_G.step()
# record
self.loss_predmeter_G.update(loss_g_all.detach())
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_G_scheduler.step()
#---------update optim_D ---------------
self.set_requires_grad(self.D, True)
self.optimizer_D.zero_grad()
temp_t = self.D(target)
temp_g = self.D(output_predframe_G.detach())
loss_d_1 = self.GANLoss(temp_t, True)
loss_d_2 = self.GANLoss(temp_g, False)
loss_d = (loss_d_1 + loss_d_2) * 0.5
loss_d.backward()
self.optimizer_D.step()
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_D_scheduler.step()
self.loss_predmeter_D.update(loss_d.detach())
# ======================End==================
self.batch_time.update(time.time() - start)
if (current_step % self.steps.param['log'] == 0):
msg = make_info_message(
current_step, self.steps.param['max'],
self.kwargs['model_type'], self.batch_time,
self.config.TRAIN.batch_size, self.data_time,
[self.loss_predmeter_G, self.loss_predmeter_D])
self.logger.info(msg)
writer.add_scalar('Train_loss_G', self.loss_predmeter_G.val,
global_steps)
writer.add_scalar('Train_loss_D', self.loss_predmeter_D.val,
global_steps)
if (current_step % self.steps.param['vis'] == 0):
vis_objects = OrderedDict({
'train_target_flow':
gtFlow_vis.detach(),
'train_pred_flow':
predFlow_vis.detach(),
'train_target_frame':
target.detach(),
'train_output_predframe_G':
output_predframe_G.detach()
})
tensorboard_vis_images(vis_objects, writer, global_steps,
self.normalize.param['train'])
global_steps += 1
# reset start
start = time.time()
self.saved_model = {'G': self.G, 'D': self.D}
self.saved_optimizer = {
'optim_G': self.optimizer_G,
'optim_D': self.optimizer_D
}
self.saved_loss = {
'loss_G': self.loss_predmeter_G.val,
'loss_D': self.loss_predmeter_D.val
}
self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])] = global_steps
def train(self, current_step):
# Pytorch [N, C, D, H, W]
# initialize
start = time.time()
self.set_requires_grad(self.F, False)
self.set_requires_grad(self.D, True)
self.set_requires_grad(self.G, True)
self.G.train()
self.D.train()
self.F.eval()
writer = self.kwargs['writer_dict']['writer']
global_steps = self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])]
# get the data
data, anno, meta = next(self._train_loader_iter)
self.data_time.update(time.time() - start)
# base on the D to get each frame
# in this method, D = 2 and not change
input_data = data[:, :, 0, :, :].cuda() # input(1-st) frame
target = data[:, :, 1, :, :].cuda() # target(2-nd) frame
# True Process =================Start===================
#---------update optim_G ---------
self.set_requires_grad(self.D, False)
output_flow_G, output_frame_G = self.G(input_data)
gt_flow_esti_tensor = torch.cat([input_data, target], 1)
flow_gt_vis, flow_gt = flow_batch_estimate(
self.F,
gt_flow_esti_tensor,
self.normalize.param['train'],
optical_size=self.config.DATASET.optical_size,
output_format=self.config.DATASET.optical_format)
fake_g = self.D(torch.cat([target, output_flow_G], dim=1))
loss_g_adv = self.GANLoss(fake_g, True)
loss_op = self.OpticalflowSqrtLoss(output_flow_G, flow_gt)
loss_int = self.IntentsityLoss(output_frame_G, target)
loss_gd = self.GradientLoss(output_frame_G, target)
loss_g_all = self.loss_lamada[
'IntentsityLoss'] * loss_int + self.loss_lamada[
'GradientLoss'] * loss_gd + self.loss_lamada[
'OpticalflowSqrtLoss'] * loss_op + self.loss_lamada[
'GANLoss'] * loss_g_adv
self.optimizer_G.zero_grad()
loss_g_all.backward()
self.optimizer_G.step()
self.loss_meter_G.update(loss_g_all.detach())
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_G_scheduler.step()
#---------update optim_D ---------------
self.set_requires_grad(self.D, True)
self.optimizer_D.zero_grad()
# import ipdb; ipdb.set_trace()
real_d = self.D(torch.cat([target, flow_gt], dim=1))
fake_d = self.D(torch.cat([target, output_flow_G.detach()], dim=1))
loss_d_1 = self.GANLoss(real_d, True)
loss_d_2 = self.GANLoss(fake_d, False)
loss_d = (loss_d_1 + loss_d_2) * 0.5
loss_d.backward()
self.optimizer_D.step()
if self.config.TRAIN.adversarial.scheduler.use:
self.optimizer_D_scheduler.step()
self.loss_meter_D.update(loss_d.detach())
# ======================End==================
self.batch_time.update(time.time() - start)
if (current_step % self.steps.param['log'] == 0):
msg = make_info_message(current_step, self.steps.param['max'],
self.kwargs['model_type'], self.batch_time,
self.config.TRAIN.batch_size,
self.data_time,
[self.loss_meter_G, self.loss_meter_D])
logger.info(msg)
writer.add_scalar('Train_loss_G', self.loss_meter_G.val, global_steps)
writer.add_scalar('Train_loss_D', self.loss_meter_D.val, global_steps)
if (current_step % self.steps.param['vis'] == 0):
temp = vis_optical_flow(
output_flow_G.detach(),
output_format=self.config.DATASET.optical_format,
output_size=(output_flow_G.shape[-2], output_flow_G.shape[-1]),
normalize=self.normalize.param['train'])
vis_objects = OrderedDict({
'train_target_flow':
flow_gt_vis.detach(),
'train_output_flow_G':
temp,
'train_target_frame':
target.detach(),
'train_output_frame_G':
output_frame_G.detach(),
})
tensorboard_vis_images(vis_objects, writer, global_steps,
self.normalize.param['train'])
global_steps += 1
# reset start
start = time.time()
# self.saved_model = {'G':self.G, 'D':self.D}
self.saved_model['G'] = self.G
self.saved_model['D'] = self.D
# self.saved_optimizer = {'optim_G': self.optimizer_G, 'optim_D': self.optimizer_D}
self.saved_optimizer['optimizer_G'] = self.optimizer_G
self.saved_optimizer['optimizer_D'] = self.optimizer_D
# self.saved_loss = {'loss_G':self.loss_meter_G.val, 'loss_D':self.loss_meter_D.val}
self.saved_loss['loss_G'] = self.loss_meter_G.val
self.saved_loss['loss_D'] = self.loss_meter_D.val
self.kwargs['writer_dict']['global_steps_{}'.format(
self.kwargs['model_type'])] = global_steps
def evaluate(self, current_step):
"""AMC evaluation method.
Evaluate the model base on some methods.
Args:
current_step: The current step at present
Returns:
results: The magnitude of the method based on this evaluation metric
"""
# Set basic things
self.engine.set_all(False)
tb_writer = self.engine.kwargs['writer_dict']['writer']
global_steps = self.engine.kwargs['writer_dict'][
'global_steps_{}'.format(self.engine.kwargs['model_type'])]
frame_num = self.engine.config.DATASET.val.sampled_clip_length
# psnr_records=[]
score_records = []
# score_records_w=[]
w_dict = OrderedDict()
# total = 0
# calc the weight for the training set
w_video_dict = self.engine.train_dataloaders_dict['w_dataset_dict']
w_video_names = self.engine.train_dataloaders_dict[
'w_dataset_dict'].keys()
for video_name in w_video_names:
# dataset = self.engine.test_dataset_dict_w[video_name]
data_loader = w_video_dict[video_name]
len_dataset = data_loader.dataset.pics_len
test_iters = len_dataset - frame_num + 1
test_counter = 0
scores = [0.0 for i in range(len_dataset)]
for data, _, _ in data_loader:
input_data_test = data[:, :, 0, :, :].cuda()
target_test = data[:, :, 1, :, :].cuda()
# import ipdb; ipdb.set_trace()
output_flow_G, output_frame_G = self.engine.G(input_data_test)
gtFlowEstim = torch.cat([input_data_test, target_test], 1)
gtFlow_vis, gtFlow = flow_batch_estimate(
self.engine.F,
gtFlowEstim,
self.engine.normalize.param['val'],
output_format=self.engine.config.DATASET.optical_format,
optical_size=self.engine.config.DATASET.optical_size)
# import ipdb; ipdb.set_trace()
diff_appe, diff_flow = simple_diff(target_test, output_frame_G,
gtFlow, output_flow_G)
# patch_score_appe, patch_score_flow, _, _ = find_max_patch(diff_appe, diff_flow)
patch_score_appe, patch_score_flow = find_max_patch(
diff_appe, diff_flow)
scores[test_counter + frame_num -
1] = [patch_score_appe, patch_score_flow]
test_counter += 1
# print(test_counter)
if test_counter >= test_iters:
scores[:frame_num - 1] = [scores[frame_num - 1]]
scores = torch.tensor(scores)
frame_w = torch.mean(scores[:, 0])
flow_w = torch.mean(scores[:, 1])
w_dict[video_name] = [len_dataset, frame_w, flow_w]
logger.info(
f'Finish calc the scores of training set {video_name} in step:{current_step}'
)
break
# import ipdb; ipdb.set_trace()
wf, wi = calc_w(w_dict)
# wf , wi = 1.0, 1.0
tb_writer.add_text('weight of train set',
f'w_f:{wf:.3f}, w_i:{wi:.3f}', global_steps)
logger.info(f'wf:{wf}, wi:{wi}')
# calc the score for the test dataset
num_videos = 0
random_video_sn = torch.randint(0, len(self.engine.test_dataset_keys),
(1, ))
for sn, video_name in enumerate(self.engine.test_dataset_keys):
num_videos += 1
# need to improve
dataloader = self.engine.val_dataloaders_dict[
'general_dataset_dict'][video_name]
len_dataset = dataloader.dataset.pics_len
test_iters = len_dataset - frame_num + 1
test_counter = 0
vis_range = range(int(len_dataset * 0.5),
int(len_dataset * 0.5 + 5))
# psnrs = np.empty(shape=(len_dataset,),dtype=np.float32)
scores = np.empty(shape=(len_dataset, ), dtype=np.float32)
for frame_sn, (data, anno, meta) in enumerate(dataloader):
test_input = data[:, :, 0, :, :].cuda()
test_target = data[:, :, 1, :, :].cuda()
g_output_flow, g_output_frame = self.engine.G(test_input)
gt_flow_esti_tensor = torch.cat([test_input, test_target], 1)
flow_gt_vis, flow_gt = flow_batch_estimate(
self.engine.F,
gt_flow_esti_tensor,
self.engine.param['val'],
output_format=self.engine.config.DATASET.optical_format,
optical_size=self.engine.config.DATASET.optical_size)
# test_psnr = psnr_error(g_output_frame, test_target)
score, _, _ = amc_score(test_target, g_output_frame, flow_gt,
g_output_flow, wf, wi)
# test_psnr = test_psnr.tolist()
score = score.tolist()
# psnrs[test_counter+frame_num-1]=test_psnr
scores[test_counter + frame_num - 1] = score
test_counter += 1
if sn == random_video_sn and (frame_sn in vis_range):
temp = vis_optical_flow(
g_output_flow.detach(),
output_format=self.engine.config.DATASET.
optical_format,
output_size=(g_output_flow.shape[-2],
g_output_flow.shape[-1]),
normalize=self.engine.normalize.param['val'])
vis_objects = OrderedDict({
'amc_eval_frame':
test_target.detach(),
'amc_eval_frame_hat':
g_output_frame.detach(),
'amc_eval_flow':
flow_gt_vis.detach(),
'amc_eval_flow_hat':
temp
})
tensorboard_vis_images(
vis_objects,
tb_writer,
global_steps,
normalize=self.engine.normalize.param['val'])
if test_counter >= test_iters:
# psnrs[:frame_num-1]=psnrs[frame_num-1]
# import ipdb; ipdb.set_trace()
scores[:frame_num - 1] = (
scores[frame_num - 1],
) # fix the bug: TypeError: can only assign an iterable
smax = max(scores)
normal_scores = np.array(
[np.divide(s, smax) for s in scores])
normal_scores = np.clip(normal_scores, 0, None)
# psnr_records.append(psnrs)
score_records.append(normal_scores)
logger.info(f'Finish test video set {video_name}')
break
# Compute the metrics based on the model's results
self.engine.pkl_path = save_score_results(
score_records,
self.engine.config,
self.engine.logger,
verbose=self.engine.verbose,
config_name=self.engine.config_name,
current_step=current_step,
time_stamp=self.engine.kwargs["time_stamp"])
results = self.engine.evaluate_function.compute(
{'val': self.engine.pkl_path})
self.engine.logger.info(results)
# Write the metric into the tensorboard
tb_writer.add_text(
f'{self.engine.config.MODEL.name}: AUC of ROC curve',
f'auc is {results.auc}', global_steps)
return results.auc
