def forward(self, x):
#x = self.dropout1(self.relu1(self.fc1(x)))
#x = self.dropout2(self.relu2(self.fc2(x)))
x = self.relu1(self.fc1(x))
x = self.relu2(self.fc2(x))
x = self.fc3(x)
return x
# if __name__ == '__main__':
config = get_config()
template_mesh = Mesh(filename='./flame_model/FLAME_sample.ply')
renderer = vis_util.SMPLRenderer(faces=template_mesh.f)
if not os.path.exists(config.out_folder):
os.makedirs(config.out_folder)
if not os.path.exists(config.out_folder + '/images'):
os.mkdir(config.out_folder + '/images')
main(config, template_mesh)
id_txt='subjects_id.txt',
R=6,
transform=composed_transforms)
if need_evaluate:
resnet50 = torch.load("./resnet50.pkl")
else:
resnet50 = models.resnet50(pretrained=True)
resnet50.cuda()
resnet50.fc = Identity()
if need_evaluate:
regression = torch.load("./model.pkl")
else:
regression = Regression()
regression.cuda()
config = get_config()
flamelayer = FLAME(config)
flamelayer.requires_grad_ = False
flamelayer.cuda()
# ringnet = SingleRingnet(resnet50, regression, flamelayer)
# ringnet.apply(weight_init)
# ringnet.cuda()
optimizer_reg = torch.optim.Adam(regression.parameters(), lr=learning_rate)
optimizer_res = torch.optim.Adam(resnet50.parameters(), lr=learning_rate)
scheduler_reg = torch.optim.lr_scheduler.StepLR(optimizer_reg, step_size=5)
scheduler_res = torch.optim.lr_scheduler.StepLR(optimizer_res, step_size=5)
# img = dataset[0]['images'][0].permute(1,2,0).numpy()
# img = dataset[0]['images'][0].permute(1,2,0)