Added Align and Order Options

.gitignore

@@ -0,0 +1,7 @@
+__pycache__
+
+*.mp4
+
+aligned_images/
+
+flagged/

app.py

@@ -2,8 +2,10 @@ import gradio as gr
 from datetime import datetime
 
 from src.face_morp import morph
+from src.utils.align_images import align_images
+from src.utils.sort_images import sort_images
 
-def transition(image_files, duration, fps, method, guideline):
+def transition(image_files, duration, fps, method, align_resize, order_images, guideline):
     time = datetime.now().strftime("%d.%m.%Y_%H.%M.%S")
     output_name = f"output_{time}_{fps}fps.mp4"
     
@@ -12,31 +14,42 @@ def transition(image_files, duration, fps, method, guideline):
     debug_messages = []
 
     try:
-        # Apelează funcția morph și prinde mesajele de debug
+        # Align and resize images
+        if align_resize:
+            aligned_dir = "aligned_images"
+            image_files =  align_images(image_files, aligned_dir)
+
+        # Sort images by age
+        if order_images:
+            image_files = sort_images(image_files)
+            
         morph(image_files, duration, fps, output_name, guideline, is_dlib)
         debug_messages.append("Video generation successful")
         return output_name, "\n".join(debug_messages)
     
     except Exception as e:
         error_message = f"Error: {str(e)}"
+        print(error_message)
         debug_messages.append(error_message)
         return None, "\n".join(debug_messages)
 
 if __name__ == "__main__":
-
     gr.Interface(
         fn=transition,
         inputs=[
             gr.File(file_count="multiple", type="filepath"),
             gr.Slider(label="Duration (seconds) between images", minimum=1, maximum=10, step=1, value=3),
-            gr.Slider(label="Frames per second (fps)", minimum=1, maximum=60, step=1, value=30),
+            gr.Slider(label="Frames per second (fps)", minimum=2, maximum=60, step=1, value=30),
             gr.Dropdown(label="Landmarks detection method", choices=["Dlib", "MediaPipe"], value="Dlib"),
+            gr.Checkbox(label="Align and Resize Images", value=True),
+            gr.Checkbox(label="Order Images by Age"),
             gr.Checkbox(label="Guideline")
+
         ],
-        outputs=[gr.Video(), gr.Textbox(label="Debug Messages")],
+        outputs=[gr.Video(), gr.Textbox(label="Output Message")],
         examples=[
-            [["examples/1.png", "examples/2.png", "examples/3.png"], 3, 30, "Dlib", False]
+            [["examples/1.png", "examples/2.png", "examples/3.png"], 3, 30, "Dlib", False, False, False]
         ],
         title="Face Morphing",
         description="Upload multiple images containing faces to create a transition video between them."
-    ).launch()
+    ).launch(share=False)

requirements.txt

@@ -4,4 +4,5 @@ numpy==1.26.4
 scipy==1.13.0
 mediapipe==0.10.11
 dlib==19.24.4
-tqdm==4.66.4
+tqdm==4.66.4
+transformers==4.40.2

src/landmark_detector.py

@@ -13,6 +13,26 @@ def read_image(image_path):
     
     return cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
 
+class LandmarksDetector:
+    def __init__(self, predictor_model_path=f'{os.path.dirname(os.path.abspath(__file__))}/utils/shape_predictor_68_face_landmarks.dat'):
+        """
+        :param predictor_model_path: path to shape_predictor_68_face_landmarks.dat file
+        """
+        self.detector = dlib.get_frontal_face_detector() # cnn_face_detection_model_v1 also can be used
+        self.shape_predictor = dlib.shape_predictor(predictor_model_path)
+
+    def get_landmarks(self, image):
+        img = dlib.load_rgb_image(image)
+        dets = self.detector(img, 1)
+        dets = [dets[0]]
+
+        for detection in dets:
+            try:
+                face_landmarks = [(item.x, item.y) for item in self.shape_predictor(img, detection).parts()]
+                yield face_landmarks
+            except:
+                print("Exception in get_landmarks()!")
+
 
 class DlibLandmarkDetector:
     def __init__(self, predictor_model_path=f'{os.path.dirname(os.path.abspath(__file__))}/utils/shape_predictor_68_face_landmarks.dat'):

src/process_images.py

@@ -18,6 +18,7 @@ def get_images_and_landmarks(image_list, is_dlib):
         raise ValueError("At least two images are required for morphing.")
         # exit()
 
+
     landmarks_list = [] # List of landmarks for each image
     images_list = []    # List of images
 
@@ -40,4 +41,9 @@ def get_images_and_landmarks(image_list, is_dlib):
         raise ValueError("At least two faces are required for morphing.")
         # exit()
 
+    # if images dont have the same dimensions raise an error
+    if len(set([image.shape for image in images_list])) > 1:
+        raise ValueError("Images must have the same dimensions for morphing.")
+        # exit()
+
     return images_list, landmarks_list

src/utils/align_images.py

@@ -0,0 +1,47 @@
+import os
+
+import argparse
+from src.utils.face_alignment import image_align
+from src.landmark_detector import LandmarksDetector
+
+
+def align_images(img_files, aligned_dir, output_size=1024, x_scale=1, y_scale=1, em_scale=0.1, use_alpha=False):
+    """
+    Extracts and aligns all faces from images using DLib and a function from original FFHQ dataset preparation step
+    python align_images.py /raw_images /aligned_images
+    """
+
+    ALIGNED_IMAGES_DIR = aligned_dir
+
+    # Create the directory if it doesn't exist
+    if not os.path.exists(ALIGNED_IMAGES_DIR):
+        os.makedirs(ALIGNED_IMAGES_DIR)
+    else:   # Remove existing files in the directory
+        for file in os.listdir(ALIGNED_IMAGES_DIR):
+            os.remove(os.path.join(ALIGNED_IMAGES_DIR, file))
+
+    landmarks_detector = LandmarksDetector()
+    for img_path in img_files:
+        img_name = os.path.basename(img_path)
+        print('Aligning %s ...' % img_name)
+        try:
+            raw_img_path = img_path
+            fn = face_img_name = '%s_%02d.png' % (os.path.splitext(img_name)[0], 1)
+            if os.path.isfile(fn):
+                continue
+            print('Getting landmarks...')
+            
+            for i, face_landmarks in enumerate(landmarks_detector.get_landmarks(raw_img_path), start=1):
+                try:
+                    print('Starting face alignment...')
+                    face_img_name = '%s_%02d.png' % (os.path.splitext(img_name)[0], i)
+                    aligned_face_path = os.path.join(ALIGNED_IMAGES_DIR, face_img_name)
+                    image_align(raw_img_path, aligned_face_path, face_landmarks, output_size=output_size, x_scale=x_scale, y_scale=y_scale, em_scale=em_scale, alpha=use_alpha)
+                    print('Wrote result %s\n' % aligned_face_path)
+                except Exception as e:
+                    raise Exception("Exception in face alignment!", e)
+        except:
+            raise Exception("Exception in landmark detection!")
+        
+    # return absolute paths of aligned images
+    return [os.path.join(ALIGNED_IMAGES_DIR, img) for img in os.listdir(ALIGNED_IMAGES_DIR)]

src/utils/face_alignment.py

@@ -0,0 +1,93 @@
+import numpy as np
+import scipy.ndimage
+import os
+import PIL.Image
+from PIL import Image
+
+
+def image_align(src_file, dst_file, face_landmarks, output_size=1024, transform_size=4096, enable_padding=True, x_scale=1, y_scale=1, em_scale=0.1, alpha=False):
+        # 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
+
+        # Load in-the-wild image.
+        if not os.path.isfile(src_file):
+            print('\nCannot find source image. Please run "--wilds" before "--align".')
+            return
+        img = PIL.Image.open(src_file).convert('RGBA').convert('RGB')
+
+        # 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, Image.Resampling.LANCZOS)
+            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))
+            if alpha:
+                mask = 1-np.clip(3.0 * mask, 0.0, 1.0)
+                mask = np.uint8(np.clip(np.rint(mask*255), 0, 255))
+                img = np.concatenate((img, mask), axis=2)
+                img = PIL.Image.fromarray(img, 'RGBA')
+            else:
+                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), Image.Resampling.LANCZOS)
+
+        # Save aligned image.
+        img.save(dst_file, 'PNG')

src/utils/sort_images.py

@@ -0,0 +1,22 @@
+from transformers import pipeline
+
+
+def sort_images(image_files):
+    pipe = pipeline("image-classification", model="Robys01/facial_age_estimator")
+
+    def get_age(image):
+        result = pipe(image)
+        print(image, "age:", result[0]["label"])
+        return result[0]["label"]
+    
+
+    image_files.sort(key=get_age)
+
+    return image_files
+
+if __name__ == "__main__":
+    image_files = ["examples/3.png", "examples/1.png", "examples/2.png"]
+    sorted_images = sort_images(image_files)
+
+
+    print(sorted_images)