lambda x: tf.assign(
x,
tf.contrib.framework.load_variable(config.load_model_dir, x.
name)), vars_list))
sess.run(assign_ops)
return output, input_image_tf, input_mask_tf
command_inputs = {
'image':
runway.image,
'mask':
runway.segmentation(label_to_id={
'background': 0,
'mask': 1
},
label_to_color={
'background': [0, 0, 0],
'mask': [255, 255, 255]
})
}
@runway.command('inpaint',
inputs=command_inputs,
outputs={'output': runway.image})
def inpaint(model, inputs):
output, input_image_tf, input_mask_tf = model
image = inputs['image']
original_size = image.size
image = np.array(image.resize((256, 256)), dtype=np.float32)
mask = np.array(inputs['mask'].resize((256, 256)), dtype=np.float32)
model.eval()
model.to(device)
print("Model:", CONFIG.MODEL.NAME)
return Dict({
'model': model,
'device': device,
'config': CONFIG,
'postprocessor': postprocessor
})
@runway.command(
'mask_all',
inputs={'image': runway.image},
outputs={'image': runway.segmentation(label_to_id=label_to_id)})
def mask_all(model, inputs):
labelmap = run_model(model, inputs).astype(np.uint8)
return {'image': labelmap}
@runway.command('mask_one',
inputs={
'image': runway.image,
'class': runway.category(choices=classes_list)
},
outputs={'image': runway.image})
def mask_one(model, inputs):
labelmap = run_model(model, inputs)
labelmap = 255.0 * np.array(
labelmap == classes_list.index(inputs['class']))
"ear_r": [215, 175, 125],
"eye_g": [220, 180, 210],
"neck_l": [125, 125, 255]
}
to_tensor = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)),
])
@runway.command('parse',
inputs={'image': runway.image},
outputs={
'parsed_face':
runway.segmentation(label_to_id=label_to_id,
label_to_color=label_to_color)
})
def parse(model, inputs):
image = inputs['image'].resize((512, 512), Image.BILINEAR)
img = to_tensor(image)
img = torch.unsqueeze(img, 0)
img = img.cuda()
out = model(img)[0]
parsing = out.squeeze(0).cpu().detach().numpy().argmax(0)
return parsing.astype(np.uint8)
if __name__ == '__main__':
runway.run(model_options={'checkpoint': './79999_iter.pth'})
label_to_color = {
'unlabeled': (0, 0, 0),
'grass': (29, 195, 49),
'sky': (95, 219, 255),
'clouds': (170, 170, 170),
'sea': (54, 62, 167),
'river': (0, 57, 150),
'tree': (140, 104, 47),
'mountain': (60, 55, 50)
}
command_inputs = {
'semantic_map':
runway.segmentation(label_to_id=label_to_id,
label_to_color=label_to_color,
default_label='unlabeled',
width=640,
height=360)
}
command_outputs = {'output': runway.image}
@runway.command('convert', inputs=command_inputs, outputs=command_outputs)
def convert(model, inputs):
img = inputs['semantic_map']
original_size = img.size
img = img.resize((opt.load_size, opt.load_size))
params = get_params(opt, img.size)
transform_label = get_transform(opt,
params,
model.eval()
return model
label_to_color = {
'unlabeled': (0, 0, 0),
'grass': (29, 195, 49),
'sky-other': (95, 219, 255),
'clouds': (170, 170, 170),
'sea': (54, 62, 167),
'river': (0, 57, 150),
'tree': (140, 104, 47),
'mountain': (60, 55, 50)
}
command_inputs = {
'semantic_map': runway.segmentation(label_to_id=label_to_id, label_to_color=label_to_color, default_label='sky-other', width=640, height=360)
}
command_outputs = {
'output': runway.image
}
@runway.command('convert', inputs=command_inputs, outputs=command_outputs)
def convert(model, inputs):
img = inputs['semantic_map']
original_size = img.size
img = img.resize((opt.load_size, opt.load_size))
params = get_params(opt, img.size)
transform_label = get_transform(opt, params, method=Image.NEAREST, normalize=False)
label_tensor = transform_label(img) * 255.0
label_tensor[label_tensor == 255.0] = 0
]
label_to_id = {label: i for i, label in enumerate(labels)}
@runway.setup(options={'checkpoint_dir': runway.file(is_directory=True)})
def setup(opts):
return DeepLabModel(opts['checkpoint_dir'])
@runway.command('segment',
inputs={'image': runway.image},
outputs={
'segmentation':
runway.segmentation(
label_to_id=label_to_id,
label_to_color={'background': [0, 0, 0]})
})
def segment(model, inputs):
_, seg_map = model.run(inputs['image'])
return seg_map.astype(np.uint8)
@runway.command('mask_person',
inputs={'image': runway.image},
outputs={'masked_image': runway.image})
def mask(model, inputs):
_, seg_map = model.run(inputs['image'])
mask = np.stack((seg_map, ) * 4, axis=-1)
masked = np.array(inputs['image'].resize(seg_map.shape[::-1]))
masked = np.dstack((masked, np.full(masked.shape[:-1], 255)))
"mouth": [255, 150, 0],
"neck": [255, 225, 120],
"r_ear": [255, 125, 125],
"l_ear": [200, 100, 100],
"cloth": [0, 255, 0],
"hat": [0, 150, 80],
"ear_r": [215, 175, 125],
"eye_g": [220, 180, 210],
"neck_l": [125, 125, 255]
}
command_inputs = {
'semantic_map':
runway.segmentation(label_to_id=label_to_id,
label_to_color=label_to_color,
default_label='background',
width=256,
height=256),
'style_image':
runway.image
}
command_outputs = {'output': runway.image}
@runway.command("generate_face",
inputs=command_inputs,
outputs=command_outputs,
description="Generates a face using SPADE")
def generate_face(sess_out, inputs):
original_size = inputs['semantic_map'].size