DeepFakeAI/processors/frame/modules/face_swapper.py

from typing import Any, List, Literal, Optional
from argparse import ArgumentParser
import threading
import numpy
import onnx
import onnxruntime
from onnx import numpy_helper

import DeepFakeAI.globals
import DeepFakeAI.processors.frame.core as frame_processors
from DeepFakeAI import logger, wording
from DeepFakeAI.face_analyser import get_one_face, get_average_face, get_many_faces, find_similar_faces, clear_face_analyser
from DeepFakeAI.face_helper import warp_face, paste_back
from DeepFakeAI.face_store import get_reference_faces
from DeepFakeAI.content_analyser import clear_content_analyser
from DeepFakeAI.typing import Face, FaceSet, Frame, Update_Process, ProcessMode, ModelSet, OptionsWithModel, Embedding
from DeepFakeAI.filesystem import is_file, is_image, are_images, is_video, resolve_relative_path
from DeepFakeAI.download import conditional_download, is_download_done
from DeepFakeAI.vision import read_image, read_static_image, read_static_images, write_image
from DeepFakeAI.processors.frame import globals as frame_processors_globals
from DeepFakeAI.processors.frame import choices as frame_processors_choices
from DeepFakeAI.face_masker import create_static_box_mask, create_occlusion_mask, create_region_mask, clear_face_occluder, clear_face_parser

FRAME_PROCESSOR = None
MODEL_MATRIX = None
THREAD_LOCK : threading.Lock = threading.Lock()
NAME = __name__.upper()
MODELS : ModelSet =\
{
	'blendswap_256':
	{
		'type': 'blendswap',
		'url': 'https://github.com/DeepFakeAI/DeepFakeAI-assets/releases/download/models/blendswap_256.onnx',
		'path': resolve_relative_path('../.assets/models/blendswap_256.onnx'),
		'template': 'ffhq_512',
		'size': (512, 256),
		'mean': [ 0.0, 0.0, 0.0 ],
		'standard_deviation': [ 1.0, 1.0, 1.0 ]
	},
	'inswapper_128':
	{
		'type': 'inswapper',
		'url': 'https://github.com/DeepFakeAI/DeepFakeAI-assets/releases/download/models/inswapper_128.onnx',
		'path': resolve_relative_path('../.assets/models/inswapper_128.onnx'),
		'template': 'arcface_128_v2',
		'size': (128, 128),
		'mean': [ 0.0, 0.0, 0.0 ],
		'standard_deviation': [ 1.0, 1.0, 1.0 ]
	},
	'inswapper_128_fp16':
	{
		'type': 'inswapper',
		'url': 'https://github.com/DeepFakeAI/DeepFakeAI-assets/releases/download/models/inswapper_128_fp16.onnx',
		'path': resolve_relative_path('../.assets/models/inswapper_128_fp16.onnx'),
		'template': 'arcface_128_v2',
		'size': (128, 128),
		'mean': [ 0.0, 0.0, 0.0 ],
		'standard_deviation': [ 1.0, 1.0, 1.0 ]
	},
	'simswap_256':
	{
		'type': 'simswap',
		'url': 'https://github.com/DeepFakeAI/DeepFakeAI-assets/releases/download/models/simswap_256.onnx',
		'path': resolve_relative_path('../.assets/models/simswap_256.onnx'),
		'template': 'arcface_112_v1',
		'size': (112, 256),
		'mean': [ 0.485, 0.456, 0.406 ],
		'standard_deviation': [ 0.229, 0.224, 0.225 ]
	},
	'simswap_512_unofficial':
	{
		'type': 'simswap',
		'url': 'https://github.com/DeepFakeAI/DeepFakeAI-assets/releases/download/models/simswap_512_unofficial.onnx',
		'path': resolve_relative_path('../.assets/models/simswap_512_unofficial.onnx'),
		'template': 'arcface_112_v1',
		'size': (112, 512),
		'mean': [ 0.0, 0.0, 0.0 ],
		'standard_deviation': [ 1.0, 1.0, 1.0 ]
	}
}
OPTIONS : Optional[OptionsWithModel] = None


def get_frame_processor() -> Any:
	global FRAME_PROCESSOR

	with THREAD_LOCK:
		if FRAME_PROCESSOR is None:
			model_path = get_options('model').get('path')
			FRAME_PROCESSOR = onnxruntime.InferenceSession(model_path, providers = DeepFakeAI.globals.execution_providers)
	return FRAME_PROCESSOR


def clear_frame_processor() -> None:
	global FRAME_PROCESSOR

	FRAME_PROCESSOR = None


def get_model_matrix() -> Any:
	global MODEL_MATRIX

	with THREAD_LOCK:
		if MODEL_MATRIX is None:
			model_path = get_options('model').get('path')
			model = onnx.load(model_path)
			MODEL_MATRIX = numpy_helper.to_array(model.graph.initializer[-1])
	return MODEL_MATRIX


def clear_model_matrix() -> None:
	global MODEL_MATRIX

	MODEL_MATRIX = None


def get_options(key : Literal['model']) -> Any:
	global OPTIONS

	if OPTIONS is None:
		OPTIONS =\
		{
			'model': MODELS[frame_processors_globals.face_swapper_model]
		}
	return OPTIONS.get(key)


def set_options(key : Literal['model'], value : Any) -> None:
	global OPTIONS

	OPTIONS[key] = value


def register_args(program : ArgumentParser) -> None:
	program.add_argument('--face-swapper-model', help = wording.get('frame_processor_model_help'), default = 'inswapper_128', choices = frame_processors_choices.face_swapper_models)


def apply_args(program : ArgumentParser) -> None:
	args = program.parse_args()
	frame_processors_globals.face_swapper_model = args.face_swapper_model
	if args.face_swapper_model == 'blendswap_256':
		DeepFakeAI.globals.face_recognizer_model = 'arcface_blendswap'
	if args.face_swapper_model == 'inswapper_128' or args.face_swapper_model == 'inswapper_128_fp16':
		DeepFakeAI.globals.face_recognizer_model = 'arcface_inswapper'
	if args.face_swapper_model == 'simswap_256' or args.face_swapper_model == 'simswap_512_unofficial':
		DeepFakeAI.globals.face_recognizer_model = 'arcface_simswap'


def pre_check() -> bool:
	if not DeepFakeAI.globals.skip_download:
		download_directory_path = resolve_relative_path('../.assets/models')
		model_url = get_options('model').get('url')
		conditional_download(download_directory_path, [ model_url ])
	return True


def pre_process(mode : ProcessMode) -> bool:
	model_url = get_options('model').get('url')
	model_path = get_options('model').get('path')
	if not DeepFakeAI.globals.skip_download and not is_download_done(model_url, model_path):
		logger.error(wording.get('model_download_not_done') + wording.get('exclamation_mark'), NAME)
		return False
	elif not is_file(model_path):
		logger.error(wording.get('model_file_not_present') + wording.get('exclamation_mark'), NAME)
		return False
	if not are_images(DeepFakeAI.globals.source_paths):
		logger.error(wording.get('select_image_source') + wording.get('exclamation_mark'), NAME)
		return False
	for source_frame in read_static_images(DeepFakeAI.globals.source_paths):
		if not get_one_face(source_frame):
			logger.error(wording.get('no_source_face_detected') + wording.get('exclamation_mark'), NAME)
			return False
	if mode in [ 'output', 'preview' ] and not is_image(DeepFakeAI.globals.target_path) and not is_video(DeepFakeAI.globals.target_path):
		logger.error(wording.get('select_image_or_video_target') + wording.get('exclamation_mark'), NAME)
		return False
	if mode == 'output' and not DeepFakeAI.globals.output_path:
		logger.error(wording.get('select_file_or_directory_output') + wording.get('exclamation_mark'), NAME)
		return False
	return True


def post_process() -> None:
	clear_frame_processor()
	clear_model_matrix()
	clear_face_analyser()
	clear_content_analyser()
	clear_face_occluder()
	clear_face_parser()
	read_static_image.cache_clear()


def swap_face(source_face : Face, target_face : Face, temp_frame : Frame) -> Frame:
	frame_processor = get_frame_processor()
	model_template = get_options('model').get('template')
	model_size = get_options('model').get('size')
	model_type = get_options('model').get('type')
	crop_frame, affine_matrix = warp_face(temp_frame, target_face.kps, model_template, model_size)
	crop_mask_list = []
	if 'box' in DeepFakeAI.globals.face_mask_types:
		crop_mask_list.append(create_static_box_mask(crop_frame.shape[:2][::-1], DeepFakeAI.globals.face_mask_blur, DeepFakeAI.globals.face_mask_padding))
	if 'occlusion' in DeepFakeAI.globals.face_mask_types:
		crop_mask_list.append(create_occlusion_mask(crop_frame))
	crop_frame = prepare_crop_frame(crop_frame)
	frame_processor_inputs = {}
	for frame_processor_input in frame_processor.get_inputs():
		if frame_processor_input.name == 'source':
			if model_type == 'blendswap':
				frame_processor_inputs[frame_processor_input.name] = prepare_source_frame(source_face)
			else:
				frame_processor_inputs[frame_processor_input.name] = prepare_source_embedding(source_face)
		if frame_processor_input.name == 'target':
			frame_processor_inputs[frame_processor_input.name] = crop_frame
	crop_frame = frame_processor.run(None, frame_processor_inputs)[0][0]
	crop_frame = normalize_crop_frame(crop_frame)
	if 'region' in DeepFakeAI.globals.face_mask_types:
		crop_mask_list.append(create_region_mask(crop_frame, DeepFakeAI.globals.face_mask_regions))
	crop_mask = numpy.minimum.reduce(crop_mask_list).clip(0, 1)
	temp_frame = paste_back(temp_frame, crop_frame, crop_mask, affine_matrix)
	return temp_frame


def prepare_source_frame(source_face : Face) -> Frame:
	source_frame = read_static_image(DeepFakeAI.globals.source_paths[0])
	source_frame, _ = warp_face(source_frame, source_face.kps, 'arcface_112_v2', (112, 112))
	source_frame = source_frame[:, :, ::-1] / 255.0
	source_frame = source_frame.transpose(2, 0, 1)
	source_frame = numpy.expand_dims(source_frame, axis = 0).astype(numpy.float32)
	return source_frame


def prepare_source_embedding(source_face : Face) -> Embedding:
	model_type = get_options('model').get('type')
	if model_type == 'inswapper':
		model_matrix = get_model_matrix()
		source_embedding = source_face.embedding.reshape((1, -1))
		source_embedding = numpy.dot(source_embedding, model_matrix) / numpy.linalg.norm(source_embedding)
	else:
		source_embedding = source_face.normed_embedding.reshape(1, -1)
	return source_embedding


def prepare_crop_frame(crop_frame : Frame) -> Frame:
	model_mean = get_options('model').get('mean')
	model_standard_deviation = get_options('model').get('standard_deviation')
	crop_frame = crop_frame[:, :, ::-1] / 255.0
	crop_frame = (crop_frame - model_mean) / model_standard_deviation
	crop_frame = crop_frame.transpose(2, 0, 1)
	crop_frame = numpy.expand_dims(crop_frame, axis = 0).astype(numpy.float32)
	return crop_frame


def normalize_crop_frame(crop_frame : Frame) -> Frame:
	crop_frame = crop_frame.transpose(1, 2, 0)
	crop_frame = (crop_frame * 255.0).round()
	crop_frame = crop_frame[:, :, ::-1].astype(numpy.uint8)
	return crop_frame


def get_reference_frame(source_face : Face, target_face : Face, temp_frame : Frame) -> Frame:
	return swap_face(source_face, target_face, temp_frame)


def process_frame(source_face : Face, reference_faces : FaceSet, temp_frame : Frame) -> Frame:
	if 'reference' in DeepFakeAI.globals.face_selector_mode:
		similar_faces = find_similar_faces(temp_frame, reference_faces, DeepFakeAI.globals.reference_face_distance)
		if similar_faces:
			for similar_face in similar_faces:
				temp_frame = swap_face(source_face, similar_face, temp_frame)
	if 'one' in DeepFakeAI.globals.face_selector_mode:
		target_face = get_one_face(temp_frame)
		if target_face:
			temp_frame = swap_face(source_face, target_face, temp_frame)
	if 'many' in DeepFakeAI.globals.face_selector_mode:
		many_faces = get_many_faces(temp_frame)
		if many_faces:
			for target_face in many_faces:
				temp_frame = swap_face(source_face, target_face, temp_frame)
	return temp_frame


def process_frames(source_paths : List[str], temp_frame_paths : List[str], update_progress : Update_Process) -> None:
	source_frames = read_static_images(source_paths)
	source_face = get_average_face(source_frames)
	reference_faces = get_reference_faces() if 'reference' in DeepFakeAI.globals.face_selector_mode else None
	for temp_frame_path in temp_frame_paths:
		temp_frame = read_image(temp_frame_path)
		result_frame = process_frame(source_face, reference_faces, temp_frame)
		write_image(temp_frame_path, result_frame)
		update_progress()


def process_image(source_paths : List[str], target_path : str, output_path : str) -> None:
	source_frames = read_static_images(source_paths)
	source_face = get_average_face(source_frames)
	reference_faces = get_reference_faces() if 'reference' in DeepFakeAI.globals.face_selector_mode else None
	target_frame = read_static_image(target_path)
	result_frame = process_frame(source_face, reference_faces, target_frame)
	write_image(output_path, result_frame)


def process_video(source_paths : List[str], temp_frame_paths : List[str]) -> None:
	frame_processors.multi_process_frames(source_paths, temp_frame_paths, process_frames)