init project

.gitignore

@@ -0,0 +1,170 @@
+#### joe made this: http://goel.io/joe
+#### macos ####
+# General
+*.DS_Store
+.AppleDouble
+.LSOverride
+
+# Icon must end with two \r
+Icon
+
+
+# Thumbnails
+._*
+
+# Files that might appear in the root of a volume
+.DocumentRevisions-V100
+.fseventsd
+.Spotlight-V100
+.TemporaryItems
+.Trashes
+.VolumeIcon.icns
+.com.apple.timemachine.donotpresent
+
+# Directories potentially created on remote AFP share
+.AppleDB
+.AppleDesktop
+Network Trash Folder
+Temporary Items
+.apdisk
+#### linux ####
+*~
+
+# temporary files which can be created if a process still has a handle open of a deleted file
+.fuse_hidden*
+
+# KDE directory preferences
+.directory
+
+# Linux trash folder which might appear on any partition or disk
+.Trash-*
+
+# .nfs files are created when an open file is removed but is still being accessed
+.nfs*
+#### windows ####
+# Windows thumbnail cache files
+Thumbs.db
+ehthumbs.db
+ehthumbs_vista.db
+
+# Dump file
+*.stackdump
+
+# Folder config file
+Desktop.ini
+
+# Recycle Bin used on file shares
+$RECYCLE.BIN/
+
+# Windows Installer files
+*.cab
+*.msi
+*.msm
+*.msp
+
+# Windows shortcuts
+*.lnk
+#### python ####
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+
+#### jetbrains ####
+
+.idea/
+.docker/
+.pytest_cache/
+*.db

.python-version

@@ -0,0 +1 @@
+miniforge3

app.py

@@ -0,0 +1,103 @@
+import argparse
+import functools
+import pathlib
+import gradio as gr
+import PIL.Image
+from encoder import Encoder
+
+from face_detector import FaceAligner
+from generator import Generator
+from huggingface_hub import hf_hub_download
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--share', action='store_true')
+    parser.add_argument('--port', type=int)
+    parser.add_argument('--disable-queue',
+                        dest='enable_queue',
+                        action='store_false')
+    return parser.parse_args()
+
+
+def load_examples():
+    image_dir = pathlib.Path('examples')
+    images = sorted(image_dir.glob('*.jpg'))
+    return [path.as_posix() for path in images]
+
+
+def predict(
+    image: PIL.Image.Image,
+    face_aligner: FaceAligner,
+    encoder: Encoder,
+    generator: Generator,
+):
+    images = face_aligner.align(image)
+
+    gen_imgs = []
+    for img in images:
+        x = encoder.predict(img)
+        gen_img = generator.predict(x)
+        gen_imgs.append(gen_img)
+
+    return gen_imgs
+
+
+def load_models():
+    encoder_path = hf_hub_download(
+        'senior-sigan/nijigenka',
+        'encoder.onnx',
+    )
+    generator_path = hf_hub_download(
+        'senior-sigan/nijigenka',
+        'face2art.onnx',
+    )
+    shape_predictor_path = hf_hub_download(
+        'senior-sigan/nijigenka',
+        'shape_predictor_68_face_landmarks.bin',
+    )
+
+    face_aligner = FaceAligner(
+        image_size=512,
+        shape_predictor_path=shape_predictor_path,
+    )
+    encoder = Encoder(model_path=encoder_path)
+    generator = Generator(model_path=generator_path)
+
+    return face_aligner, encoder, generator
+
+
+def main():
+    args = parse_args()
+    gr.close_all()
+
+    face_aligner, encoder, generator = load_models()
+
+    func = functools.partial(
+        predict,
+        face_aligner=face_aligner,
+        encoder=encoder,
+        generator=generator,
+    )
+    func = functools.update_wrapper(func, predict)
+
+    iface = gr.Interface(
+        fn=func,
+        inputs=[
+            gr.inputs.Image(type='pil', label='Input')
+        ],
+        outputs=gr.outputs.Carousel(['image']),
+        examples=load_examples(),
+        title='Nijigenka: Portrait to Art',
+        allow_flagging='never',
+        theme='huggingface',
+    )
+    iface.launch(
+        enable_queue=args.enable_queue,
+        server_port=args.port,
+        share=args.share,
+    )
+
+
+if __name__ == '__main__':
+    main()

encoder.py

@@ -0,0 +1,32 @@
+import PIL.Image
+import onnxruntime
+import numpy as np
+
+
+def to_tensor(
+    image: PIL.Image.Image,
+    image_size=(256, 256),
+) -> np.ndarray:
+    image = image.resize(image_size)
+    x = np.asarray(image) / 255.0
+    x = np.transpose(x, (2, 0, 1))
+    x = (x - 0.5) / 0.5
+    x = np.expand_dims(x, axis=0).astype(np.float32)
+    return x
+
+
+class Encoder(object):
+    def __init__(
+        self,
+        model_path: str = 'encoder.onnx',
+    ) -> None:
+        self.input_name = 'input_0'
+        self.output_name = 'output_0'
+        self.session = onnxruntime.InferenceSession(model_path, None)
+
+    def predict(self, image: PIL.Image.Image) -> np.ndarray:
+        x = to_tensor(image)
+        output = self.session.run([self.output_name], {
+            self.input_name: x,
+        })[0][0]
+        return output

face_detector.py

@@ -0,0 +1,135 @@
+from typing import Iterator, List, Tuple
+
+import dlib
+import numpy as np
+import PIL.Image
+import scipy.ndimage
+
+
+class FaceAligner(object):
+    def __init__(
+        self,
+        shape_predictor_path: str = 'shape_predictor_68_face_landmarks.dat',
+        image_size: int = 512,
+    ) -> None:
+        self.image_size = image_size
+        self.detector = dlib.get_frontal_face_detector()
+        self.shape_predictor = dlib.shape_predictor(shape_predictor_path)
+
+    def align(self, image: PIL.Image.Image) -> List[PIL.Image.Image]:
+        landmarks = self.get_landmarks(image)
+
+        return [image_align(
+            image,
+            face_landmarks,
+            output_size=self.image_size,
+            transform_size=self.image_size * 2,
+        ) for face_landmarks in landmarks]
+
+    def get_landmarks(
+        self,
+        image: PIL.Image.Image,
+    ) -> Iterator[List[Tuple[int, int]]]:
+        img = np.asarray(image.convert('L'))
+        dets = self.detector(img, 1)
+
+        for detection in dets:
+            try:
+                parts = self.shape_predictor(img, detection).parts()
+                face_landmarks = [(point.x, point.y) for point in parts]
+                yield face_landmarks
+            except:
+                print("Exception in get_landmarks()!")
+
+
+def image_align(
+    img: PIL.Image.Image,
+    face_landmarks: List[Tuple[int, int]],
+    output_size: int = 1024,
+    transform_size: int = 4096,
+    enable_padding: bool = True,
+    x_scale: float = 1,
+    y_scale: float = 1,
+    em_scale: float = 0.1,
+) -> PIL.Image.Image:
+    # Align function from FFHQ dataset pre-processing step
+    # https://github.com/NVlabs/ffhq-dataset/blob/master/download_ffhq.py
+
+    lm = np.array(face_landmarks)
+    lm_chin = lm[0: 17]  # left-right
+    lm_eyebrow_left = lm[17: 22]  # left-right
+    lm_eyebrow_right = lm[22: 27]  # left-right
+    lm_nose = lm[27: 31]  # top-down
+    lm_nostrils = lm[31: 36]  # top-down
+    lm_eye_left = lm[36: 42]  # left-clockwise
+    lm_eye_right = lm[42: 48]  # left-clockwise
+    lm_mouth_outer = lm[48: 60]  # left-clockwise
+    lm_mouth_inner = lm[60: 68]  # left-clockwise
+
+    # Calculate auxiliary vectors.
+    eye_left = np.mean(lm_eye_left, axis=0)
+    eye_right = np.mean(lm_eye_right, axis=0)
+    eye_avg = (eye_left + eye_right) * 0.5
+    eye_to_eye = eye_right - eye_left
+    mouth_left = lm_mouth_outer[0]
+    mouth_right = lm_mouth_outer[6]
+    mouth_avg = (mouth_left + mouth_right) * 0.5
+    eye_to_mouth = mouth_avg - eye_avg
+
+    # Choose oriented crop rectangle.
+    x = eye_to_eye - np.flipud(eye_to_mouth) * [-1, 1]
+    x /= np.hypot(*x)
+    x *= max(np.hypot(*eye_to_eye) * 2.0, np.hypot(*eye_to_mouth) * 1.8)
+    x *= x_scale
+    y = np.flipud(x) * [-y_scale, y_scale]
+    c = eye_avg + eye_to_mouth * em_scale
+    quad = np.stack([c - x - y, c - x + y, c + x + y, c + x - y])
+    qsize = np.hypot(*x) * 2
+
+    # Shrink.
+    shrink = int(np.floor(qsize / output_size * 0.5))
+    if shrink > 1:
+        rsize = (int(np.rint(float(img.size[0]) / shrink)),
+                 int(np.rint(float(img.size[1]) / shrink)))
+        img = img.resize(rsize, PIL.Image.ANTIALIAS)
+        quad /= shrink
+        qsize /= shrink
+
+    # Crop.
+    border = max(int(np.rint(qsize * 0.1)), 3)
+    crop = (int(np.floor(min(quad[:, 0]))), int(np.floor(min(quad[:, 1]))), int(
+        np.ceil(max(quad[:, 0]))), int(np.ceil(max(quad[:, 1]))))
+    crop = (max(crop[0] - border, 0), max(crop[1] - border, 0),
+            min(crop[2] + border, img.size[0]), min(crop[3] + border, img.size[1]))
+    if crop[2] - crop[0] < img.size[0] or crop[3] - crop[1] < img.size[1]:
+        img = img.crop(crop)
+        quad -= crop[0:2]
+
+    # Pad.
+    pad = (int(np.floor(min(quad[:, 0]))), int(np.floor(min(quad[:, 1]))), int(
+        np.ceil(max(quad[:, 0]))), int(np.ceil(max(quad[:, 1]))))
+    pad = (max(-pad[0] + border, 0), max(-pad[1] + border, 0), max(pad[2] -
+           img.size[0] + border, 0), max(pad[3] - img.size[1] + border, 0))
+    if enable_padding and max(pad) > border - 4:
+        pad = np.maximum(pad, int(np.rint(qsize * 0.3)))
+        img = np.pad(np.float32(img),
+                     ((pad[1], pad[3]), (pad[0], pad[2]), (0, 0)), 'reflect')
+        h, w, _ = img.shape
+        y, x, _ = np.ogrid[:h, :w, :1]
+        mask = np.maximum(1.0 - np.minimum(np.float32(x) / pad[0], np.float32(
+            w-1-x) / pad[2]), 1.0 - np.minimum(np.float32(y) / pad[1], np.float32(h-1-y) / pad[3]))
+        blur = qsize * 0.02
+        img += (scipy.ndimage.gaussian_filter(img,
+                                              [blur, blur, 0]) - img) * np.clip(mask * 3.0 + 1.0, 0.0, 1.0)
+        img += (np.median(img, axis=(0, 1)) - img) * np.clip(mask, 0.0, 1.0)
+        img = np.uint8(np.clip(np.rint(img), 0, 255))
+        img = PIL.Image.fromarray(img, 'RGB')
+        quad += pad[:2]
+
+    # Transform.
+    img = img.transform((transform_size, transform_size),
+                        PIL.Image.QUAD, (quad + 0.5).flatten(), PIL.Image.BILINEAR)
+    if output_size < transform_size:
+        img = img.resize((output_size, output_size), PIL.Image.ANTIALIAS)
+
+    return img

generator.py

@@ -0,0 +1,28 @@
+from xml.etree.ElementTree import PI
+import PIL.Image
+import onnxruntime
+import numpy as np
+
+
+def to_image(tensor: np.ndarray) -> PIL.Image.Image:
+    tensor = tensor * 0.5 + 0.5
+    tensor = np.clip(tensor * 255, 0, 255).astype(np.uint8)
+    tensor = np.transpose(tensor, (1, 2, 0))
+    return PIL.Image.fromarray(tensor)
+
+
+class Generator(object):
+    def __init__(
+        self,
+        model_path: str = 'generator.onnx',
+    ) -> None:
+        self.input_name = 'input_0'
+        self.output_name = 'output_0'
+        self.session = onnxruntime.InferenceSession(model_path, None)
+
+    def predict(self, x: np.ndarray) -> PIL.Image.Image:
+        x = np.expand_dims(x, 0)
+        output = self.session.run([self.output_name], {
+            self.input_name: x,
+        })[0][0]
+        return to_image(output)