examples

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text

.idea/.gitignore

@@ -0,0 +1,8 @@
+# Default ignored files
+/shelf/
+/workspace.xml
+# Editor-based HTTP Client requests
+/httpRequests/
+# Datasource local storage ignored files
+/dataSources/
+/dataSources.local.xml

.idea/facial_beauty_analysis.iml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<module type="PYTHON_MODULE" version="4">
+  <component name="NewModuleRootManager">
+    <content url="file://$MODULE_DIR$" />
+    <orderEntry type="inheritedJdk" />
+    <orderEntry type="sourceFolder" forTests="false" />
+  </component>
+</module>

.idea/inspectionProfiles/profiles_settings.xml

@@ -0,0 +1,6 @@
+<component name="InspectionProjectProfileManager">
+  <settings>
+    <option name="USE_PROJECT_PROFILE" value="false" />
+    <version value="1.0" />
+  </settings>
+</component>

.idea/misc.xml

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="Black">
+    <option name="sdkName" value="Python 3.11" />
+  </component>
+  <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.11" project-jdk-type="Python SDK" />
+</project>

.idea/modules.xml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="ProjectModuleManager">
+    <modules>
+      <module fileurl="file://$PROJECT_DIR$/.idea/facial_beauty_analysis.iml" filepath="$PROJECT_DIR$/.idea/facial_beauty_analysis.iml" />
+    </modules>
+  </component>
+</project>

.idea/vcs.xml

@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="" vcs="Git" />
+  </component>
+</project>

app.py

@@ -1,7 +1,55 @@
+import os
+
+import cv2
 import gradio as gr
+import numpy as np
+import joblib
+
+from api import is_beautiful
+from facial_image import FacialImage, convert_to_numpy_array
+
+title = "Sytoss System: Beauty Recognition"
+description = "A model to classify is face beautiful or not."
+
+target_width = 512
+facial = FacialImage()
+loaded_svm_classifier = joblib.load('svm_model.pkl')
+
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+
+def process_image(input_image: np.ndarray):
+    if input_image is not None:
+
+        output_image = input_image
+
+        input_height, input_width, _ = input_image.shape
+        if input_width > target_width:
+            scale_factor = float(target_width / input_width)
+            output_width = int(input_width * scale_factor)
+            output_height = int(input_height * scale_factor)
+            dsize = (output_width, output_height)
+            output_image = cv2.resize(input_image, dsize)
+
+        ratios, output_image = facial.calculate_ratios(output_image)
+        ratios_vector = convert_to_numpy_array(ratios)
+
+        class_probabilities = loaded_svm_classifier.predict_proba([ratios_vector])
+
+        predicted_class = np.argmax(class_probabilities)
+
+        probability_of_predicted_class = class_probabilities[0, predicted_class]
+
+        return output_image, predicted_class, probability_of_predicted_class
+    return None, None
+
 
-def greet(name):
-    return "Hello " + name + "!!"
+iface = gr.Interface(
+    process_image,
+    inputs=gr.inputs.Image(),
+    outputs=["image", gr.Number(label="Predicted class"), gr.Number(label="Probability")],
+    title=title,
+    description=description,
+    examples=example_list)
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+iface.launch()

facial_image.py

@@ -0,0 +1,92 @@
+from math import dist
+
+import cv2
+import mediapipe as mp
+
+from points import points
+from ratios import Ratios
+from utils import *
+
+
+def convert_to_numpy_array(ratios: dict[str, float]) -> numpy.array:
+    arr = []
+    for k, v in ratios.items():
+        arr.append(v)
+    return numpy.array(arr)
+
+
+class FacialImage:
+    def __init__(self):
+        self.landmarks = []
+        self.width = 0
+        self.height = 0
+
+    def get_point_coordinates(self, point_num):
+        point = self.landmarks[point_num]
+        return point.x * self.width, point.y * self.height
+
+    def get_distance_two_points(self, point_num_1, point_num_2):
+        first = self.get_point_coordinates(point_num_1)
+        second = self.get_point_coordinates(point_num_2)
+        return dist(first, second)
+
+    def get_ratio_between_two(self, ratio1_name, ratio2_name):
+        first = self.get_distance_two_points(points[ratio1_name][0], points[ratio1_name][1])
+        second = self.get_distance_two_points(points[ratio2_name][0], points[ratio2_name][1])
+        return first / second
+
+    def get_landmarks(self, image: numpy.ndarray):
+        mp_face_mesh = mp.solutions.face_mesh
+        face_mesh = mp_face_mesh.FaceMesh()
+
+        # Image
+        height, width, _ = image.shape
+        rgb_image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+
+        # Facial landmarks
+        result = face_mesh.process(rgb_image)
+
+        for facial_landmarks in result.multi_face_landmarks:
+            for i in range(0, 468):
+                pt1 = facial_landmarks.landmark[i]
+                x = int(pt1.x * width)
+                y = int(pt1.y * height)
+                cv2.circle(image, (x, y), 1, (255, 0, 0), -1)
+
+        self.landmarks = result.multi_face_landmarks[0].landmark
+        self.width = width
+        self.height = height
+
+    def calculate_ratios(self, image: numpy.ndarray, is_normalized=False):
+        self.get_landmarks(image)
+
+        ratios = {
+            'Under eyes/Interocular': self.get_ratio_between_two(Ratios.UNDER_EYES, Ratios.INTEROCULAR),
+            'Under eyes/Nose width': self.get_ratio_between_two(Ratios.UNDER_EYES, Ratios.NOSE_WIDTH),
+            'Mouth width/Interocular': self.get_ratio_between_two(Ratios.MOUTH_WIDTH, Ratios.INTEROCULAR),
+            'Upper lip-jaw/Interocular': self.get_ratio_between_two(Ratios.UPPER_LIP_TO_JAW, Ratios.INTEROCULAR),
+            'Upper lip-jaw/Nose width': self.get_ratio_between_two(Ratios.UPPER_LIP_TO_JAW, Ratios.NOSE_WIDTH),
+            'Interocular/Lip height': self.get_ratio_between_two(Ratios.INTEROCULAR, Ratios.LIPS_HEIGHT),
+            'Nose width/Interocular': self.get_ratio_between_two(Ratios.NOSE_WIDTH, Ratios.INTEROCULAR),
+            'Nose width/Upper lip height': self.get_ratio_between_two(Ratios.NOSE_WIDTH, Ratios.UPPER_LIP_HEIGHT),
+            'Interocular/Nose mouth height': self.get_ratio_between_two(Ratios.INTEROCULAR,
+                                                                        Ratios.NOSE_TO_MOUTH_HEIGHT),
+            'Face top-eyebrows/Eyebrows-Nose': self.get_ratio_between_two(Ratios.FACE_TOP_TO_EYEBROWS,
+                                                                          Ratios.EYEBROWS_TO_NOSE),
+            'Eyebrows-nose/Nose-jaw': self.get_ratio_between_two(Ratios.EYEBROWS_TO_NOSE, Ratios.NOSE_TO_JAW),
+            'Face top-eyebrows/Nose-Jaw': self.get_ratio_between_two(Ratios.FACE_TOP_TO_EYEBROWS, Ratios.NOSE_TO_JAW),
+            'Interocular/Nose width': self.get_ratio_between_two(Ratios.INTEROCULAR, Ratios.NOSE_WIDTH),
+            'Face height/Face width': self.get_ratio_between_two(Ratios.FACE_HEIGHT, Ratios.FACE_WIDTH),
+            'Lower eyebrow length': self.get_ratio_between_two(Ratios.LEFT_LOWER_EYEBROW_LENGTH,
+                                                               Ratios.RIGHT_LOWER_EYEBROW_LENGTH),
+            'Lower lip length': self.get_ratio_between_two(Ratios.LEFT_LOWER_LIP_LENGTH, Ratios.RIGHT_LOWER_LIP_LENGTH),
+            'Upper eyebrow': self.get_ratio_between_two(Ratios.LEFT_UPPER_EYEBROW_LENGTH,
+                                                        Ratios.RIGHT_UPPER_EYEBROW_LENGTH),
+            'Upper lip': self.get_ratio_between_two(Ratios.LEFT_UPPER_LIP_LENGTH, Ratios.RIGHT_UPPER_LIP_LENGTH),
+            'Nose': self.get_ratio_between_two(Ratios.LEFT_NOSE_WIDTH, Ratios.RIGHT_NOSE_WIDTH)
+        }
+
+        if is_normalized:
+            return normalization(ratios), image
+
+        return ratios, image

points.py

@@ -0,0 +1,27 @@
+from ratios import Ratios
+
+points = {
+    Ratios.UNDER_EYES: (23, 253),
+    Ratios.INTEROCULAR: (243, 463),
+    Ratios.NOSE_WIDTH: (48, 278),
+    Ratios.MOUTH_WIDTH: (61, 293),
+    Ratios.UPPER_LIP_TO_JAW: (0, 152),
+    Ratios.LIPS_HEIGHT: (0, 17),
+    Ratios.NOSE_TO_MOUTH_HEIGHT: (2, 0),
+    Ratios.UPPER_LIP_HEIGHT: (0, 13),
+    Ratios.FACE_TOP_TO_EYEBROWS: (10, 9),
+    Ratios.EYEBROWS_TO_NOSE: (9, 2),
+    Ratios.NOSE_TO_JAW: (2, 152),
+    Ratios.FACE_HEIGHT: (10, 152),
+    Ratios.FACE_WIDTH: (227, 447),
+    Ratios.LEFT_LOWER_EYEBROW_LENGTH: (46, 55),
+    Ratios.RIGHT_LOWER_EYEBROW_LENGTH: (285, 276),
+    Ratios.LEFT_LOWER_LIP_LENGTH: (61, 17),
+    Ratios.RIGHT_LOWER_LIP_LENGTH: (17, 291),
+    Ratios.LEFT_UPPER_EYEBROW_LENGTH: (105, 9),
+    Ratios.RIGHT_UPPER_EYEBROW_LENGTH: (334, 9),
+    Ratios.LEFT_UPPER_LIP_LENGTH: (61, 0),
+    Ratios.RIGHT_UPPER_LIP_LENGTH: (17, 0),
+    Ratios.LEFT_NOSE_WIDTH: (48, 2),
+    Ratios.RIGHT_NOSE_WIDTH: (278, 2)
+}

ratios.py

@@ -0,0 +1,27 @@
+import enum
+
+
+class Ratios(enum.Enum):
+    UNDER_EYES = "under_eyes"
+    INTEROCULAR = "interocular"
+    NOSE_WIDTH = "nose_width"
+    MOUTH_WIDTH = "mouth_width"
+    UPPER_LIP_TO_JAW = "upper_lip_to_jaw"
+    LIPS_HEIGHT = "lips_height"
+    NOSE_TO_MOUTH_HEIGHT = "nose_to_mouth_height"
+    UPPER_LIP_HEIGHT = "upper_lip_height"
+    FACE_TOP_TO_EYEBROWS = "face_top_to_eyebrows"
+    EYEBROWS_TO_NOSE = "eyebrows_to_nose"
+    NOSE_TO_JAW = "nose_to_jaw"
+    FACE_HEIGHT = "face_height"
+    FACE_WIDTH = "face_width"
+    LEFT_LOWER_EYEBROW_LENGTH = "left_lower_eyebrow_length"
+    RIGHT_LOWER_EYEBROW_LENGTH = "right_lower_eyebrow_length"
+    LEFT_LOWER_LIP_LENGTH = "left_lower_lip_length"
+    RIGHT_LOWER_LIP_LENGTH = "right_lower_lip_length"
+    LEFT_UPPER_EYEBROW_LENGTH = "left_upper_eyebrow_length"
+    RIGHT_UPPER_EYEBROW_LENGTH = "right_upper_eyebrow_length"
+    LEFT_UPPER_LIP_LENGTH = "left_upper_lip_length"
+    RIGHT_UPPER_LIP_LENGTH = "right_upper_lip_length"
+    LEFT_NOSE_WIDTH = "nose_left_width"
+    RIGHT_NOSE_WIDTH = "nose_right_width"

svm_model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26064833ae014765f7c01bed1704ff860a4ac16c1a3d77b39f1a1a32a8abef64
+size 354762

utils.py

@@ -0,0 +1,55 @@
+import numpy
+import csv
+
+
+def write_to_csv(ratios, file_name):
+    rows = []
+
+    for i in range(len(ratios)):
+        row = []
+        for key, value in ratios[i].items():
+            row.append(value)
+        rows.append(row)
+
+    with open(file_name, 'a', newline='') as csvfile:
+        csvwriter = csv.writer(csvfile)
+        csvwriter.writerows(rows)
+        csvfile.close()
+
+
+def write_line_to_csv(ratios, file_name):
+    row = []
+
+    for key, value in ratios.items():
+        row.append(value)
+
+    with open(file_name, 'a', newline='') as csvfile:
+        csvwriter = csv.writer(csvfile)
+        csvwriter.writerow(row)
+
+
+def standard_deviation(values):
+    _mean = numpy.mean(values)
+    differences = [(value - _mean) ** 2 for value in values]
+    sum_of_differences = sum(differences)
+    return (sum_of_differences / (len(values) - 1)) ** 0.5
+
+
+def normalization(ratios_list):
+    _mean = numpy.mean(list(ratios_list.values()))
+    _deviation = standard_deviation(list(ratios_list.values()))
+
+    z_score = {}
+    for z_key, z_i in ratios_list.items():
+        z_score[z_key] = (z_i - _mean) / _deviation
+
+    lb = 0
+    ub = 1.618
+    _min = min(z_score.values())
+    _max = max(z_score.values())
+
+    linear = {}
+    for z_key, z_i in z_score.items():
+        linear[z_key] = lb + (z_i - _min) * (ub - lb) / (_max - _min)
+
+    return linear