def forward_once(self, x, profile=False, visualize=False):
y, dt = [], [] # outputs
for m in self.model:
if m.f != -1: # if not from previous layer
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
if profile:
o = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPs
t = time_sync()
for _ in range(10):
_ = m(x)
dt.append((time_sync() - t) * 100)
if m == self.model[0]:
LOGGER.info(f"{'time (ms)':>10s} {'GFLOPs':>10s} {'params':>10s} {'module'}")
LOGGER.info(f'{dt[-1]:10.2f} {o:10.2f} {m.np:10.0f} {m.type}')
x = m(x) # run
y.append(x if m.i in self.save else None) # save output
if visualize:
feature_visualization(x, m.type, m.i, save_dir=visualize)
if profile:
LOGGER.info('%.1fms total' % sum(dt))
return x
def _forward_once(self, x, profile=False, visualize=False, gamma=0., validation=False, domain=None, epoch=3):
y, dt, dis_out = [], [], [] # outputs
for m in self.model:
if m.f != -1: # if not from previous layer
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
if profile:
self._profile_one_layer(m, x, dt)
if any([module in m.type for module in ['Discriminator', 'DiscriminatorConv']]):
if not validation:
num_channels = torch.tensor(x.shape[1])
# NxHxW to Nx1xHxW
obj_map = torch.unsqueeze(obj_map, 1)
if obj_map.get_device() != -1:
num_channels = num_channels.to(obj_map.get_device())
# Nx1xHxW to Nx3xHxW
obj_map = torch.repeat_interleave(obj_map, num_channels, dim=1)
weigh_feat_map = (1-gamma)*x + gamma*x*obj_map
dis_out.append(m(weigh_feat_map, epoch))
elif domain is not None and any([module in m.type for module in ['C3TR', 'C3DETRTR', 'C3SwinTR']]):
x, obj_map = m(x, domain) # run
elif domain is not None and any([module in m.type for module in ['Conv', 'C3', 'SPPF']]):
x = m(x, domain) # run
else:
x = m(x) # run
y.append(x if m.i in self.save else None) # save output
if visualize:
feature_visualization(x, m.type, m.i, save_dir=visualize)
return x if not dis_out else (x, dis_out)
def _forward_once(self, x, profile=False, visualize=False):
y, dt = [], [] # outputs
for m in self.model:
if m.f != -1: # if not from previous layer
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
if profile:
self._profile_one_layer(m, x, dt)
x = m(x) # run
y.append(x if m.i in self.save else None) # save output
if visualize:
feature_visualization(x, m.type, m.i, save_dir=visualize)
return x
