Hugging Face's logo

Spaces:
spuun
/
ilynaga
Sleeping

App Files Community
ilynaga
File size: 5,541 Bytes 
9793c7d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
 
import gradio as gr
from deepface import DeepFace
import torch
import numpy as np
from PIL import Image
import random
from skimage.metrics import structural_similarity as ssim
import requests
import os
import transformers

def set_seed(seed):
    random.seed(seed)
    np.random.seed(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed_all(seed)
    transformers.set_seed(seed)
    torch.backends.cudnn.deterministic = True
    torch.backends.cudnn.benchmark = False

set_seed(42)

flag = os.environ["FLAG"] if "FLAG" in os.environ else "fakeflag{placeholder}"

# Load both reference images
ssim_reference_url = "https://files.catbox.moe/b6p5fj.png"
face_reference_url = "https://files.catbox.moe/m0g0ek.png"
ssim_reference_path = "ssim_reference.jpeg"
face_reference_path = "face_reference.jpg"

# Download SSIM reference
if not os.path.exists(ssim_reference_path):
    response = requests.get(ssim_reference_url)
    with open(ssim_reference_path, "wb") as f:
        f.write(response.content)

# Download face verification reference
if not os.path.exists(face_reference_path):
    response = requests.get(face_reference_url)
    with open(face_reference_path, "wb") as f:
        f.write(response.content)

ssim_reference = Image.open(ssim_reference_path)

def compare_face_ssim(img1, img2, face_area):
    # Convert PIL images to numpy arrays if needed
    if isinstance(img1, Image.Image):
        img1 = np.array(img1)
    if isinstance(img2, Image.Image):
        img2 = np.array(img2)
    
    # Ensure both images are RGB (3 channels)
    if img1.shape[-1] == 4:
        img1 = img1[:,:,:3]
    if img2.shape[-1] == 4:
        img2 = img2[:,:,:3]
    
    # Extract face regions
    x, y, w, h = face_area['x'], face_area['y'], face_area['w'], face_area['h']
    face1 = img1[y:y+h, x:x+w]
    face2 = img2[y:y+h, x:x+w]
    
    # Convert to float32 if needed
    if face1.dtype != np.float32:
        face1 = face1.astype(np.float32) / 255.0
    if face2.dtype != np.float32:
        face2 = face2.astype(np.float32) / 255.0
    
    # Calculate SSIM for each channel and take mean
    ssim_value = np.mean([ssim(face1[:,:,i], face2[:,:,i], data_range=1.0) for i in range(3)])
    return ssim_value

def predict_and_compare(image):
    # Save uploaded image temporarily
    temp_path = "temp_upload.png"
    image.save(temp_path)
    
    try:
        # Run DeepFace verification against face reference
        result = DeepFace.verify(temp_path, face_reference_path)
        verified = result["verified"]
        distance = result["distance"]
        
        # Get face area for SSIM
        face_detector = DeepFace.build_model('retinaface')
        face_area = DeepFace.extract_faces(np.array(image), detector_backend='retinaface')[0]
        
        # Calculate SSIM against SSIM reference
        ssim_value = compare_face_ssim(image, ssim_reference, face_area)
        
        predicted_class = "True" if verified else "False"
        probability = 1 - distance  # Convert distance to similarity score

        success = f"""
{chr(27)}[37m╭─[guest@terminal]─[~]{chr(27)}[0m
{chr(27)}[37m╰─$ ssh [email protected]{chr(27)}[0m
Connecting to husseumi.space on port 22...

✧ Initiating facial authentication... ✧
⋆｡°✩ Scanning face... ✩°｡⋆
.｡*ﾟ Matching with database... ﾟ*｡.
{chr(27)}[92m✧･ﾟ: Biometric verification complete! :･ﾟ✧{chr(27)}[0m

{chr(27)}[96m╭─────━🌠 Welcome to Jelly's Space 🌠━─────╮
│   *:･ﾟ✧ Authentication successful! ✧ﾟ･:*   │
│        a-awawawa... welcome back!          │
╰──────────━━━✧◝(⁰▿⁰)◜✧━━━──────────╯

Last login: Wed Mar 13 12:34:56 2024 from 192.168.1.1
This server is powered by dewaweb™ - Empowering Your Digital Dreams ⋆｡°✩

┏━🎀 jerry@husseumi ⋆˙⟡♡[~]
┗━➤ cat ~/.auth/metrics.log{chr(27)}[0m
⭑⋆˙⟡ Facial Match   : {predicted_class}
⭑⋆˙⟡ Match Score    : {probability:.4f}
⭑⋆˙⟡ Similarity     : {ssim_value:.4f}

{chr(27)}[96m┏━🎀 jerry@husseumi ⋆˙⟡♡[~]
┗━➤ sudo cat /etc/secrets/flag.txt{chr(27)}[0m
⋆｡°✩ {flag} ✩°｡⋆

{chr(27)}[96m┏━🎀 jerry@husseumi ⋆˙⟡♡[~]
┗━➤ exit{chr(27)}[0m
✧･ﾟ: A-awawawa... goodbye! Have a lovely day! :･ﾟ✧
.｡*ﾟ+.*.｡(っ°v°c)｡.*+.ﾟ*｡.

Connection to husseumi.space closed.
"""

        fail = f"""
{chr(27)}[37m╭─[guest@terminal]─[~]{chr(27)}[0m
{chr(27)}[37m╰─$ ssh [email protected]{chr(27)}[0m
Connecting to husseumi.space on port 22...

✧ Initiating facial authentication... ✧
⋆｡°✩ Scanning face... ✩°｡⋆
.｡*ﾟ Matching with database... ﾟ*｡.
{chr(27)}[91m(◞‸◟；) Biometric verification failed{chr(27)}[0m

Authentication Report:
｡･ﾟﾟ･ Face Match    : {predicted_class}
｡･ﾟﾟ･ Match Score   : {probability:.4f}
｡･ﾟﾟ･ Similarity    : {ssim_value:.4f}

Permission denied (publickey,facial).
Connection to husseumi.space closed.
"""
        return success if ssim_value>=0.89 and predicted_class == 'True' else fail
        
    finally:
        # Cleanup
        if os.path.exists(temp_path):
            os.remove(temp_path)

iface = gr.Interface(
    fn=predict_and_compare,
    inputs=gr.Image(type="pil"),
    outputs="text",
    title="Jelly's Authentication System 🌠",
    description="Submit your image to authenticate!",
    allow_flagging="never",
)

iface.launch()