Spaces:
Sleeping
Sleeping
File size: 2,621 Bytes
8c30b41 2d38802 c03cdfb 9e3f29f b8bf0c0 9e3f29f c03cdfb 9e3f29f c03cdfb 9e3f29f c03cdfb 9e3f29f c03cdfb 9e3f29f c03cdfb 9e3f29f c03cdfb 8c30b41 9e3f29f 8c30b41 9e3f29f |
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 |
import gradio as gr
from transformers import pipeline
from PIL import Image
import numpy as np
# ----------------------------
# Inisialisasi model
# ----------------------------
# Model A: AI detector wajah
model_a = pipeline("image-classification", model="rodrigoberriel/real-or-fake-face")
# Model B: AI detector general (misal: resnet-50)
model_b = pipeline("image-classification", model="microsoft/resnet-50")
# Model C: General image classifier
model_c = pipeline("image-classification", model="google/vit-base-patch16-224")
# ----------------------------
# Fungsi prediksi dengan ensemble
# ----------------------------
def detect_image(image):
results = []
# Model A
res_a = model_a(image)[0]
label_a, score_a = res_a['label'].lower(), res_a['score']
results.append((label_a, score_a))
# Model B
res_b = model_b(image)[0]
label_b, score_b = res_b['label'].lower(), res_b['score']
# Simplifikasi: jika label mirip 'human' atau 'person' = human, lainnya = AI
if 'human' in label_b or 'person' in label_b:
label_b_final = 'human'
else:
label_b_final = 'ai'
results.append((label_b_final, score_b))
# Model C
res_c = model_c(image)[0]
label_c, score_c = res_c['label'].lower(), res_c['score']
if 'human' in label_c or 'person' in label_c:
label_c_final = 'human'
else:
label_c_final = 'ai'
results.append((label_c_final, score_c))
# ----------------------------
# Voting + Threshold
# ----------------------------
votes = [r[0] for r in results]
scores = [r[1] for r in results]
# Mayoritas voting
final_label = max(set(votes), key=votes.count)
# Rata-rata confidence untuk final label
relevant_scores = [s for (l, s) in results if l == final_label]
avg_confidence = np.mean(relevant_scores) * 100 # ke persen
# Tentukan hasil akhir
if avg_confidence < 80:
return "⚠️ Tidak Pasti (cek manual)", round(avg_confidence, 2)
elif final_label == 'human':
return f"Foto Asli ✅", round(avg_confidence, 2)
else:
return f"AI Detected 🤖", round(avg_confidence, 2)
# ----------------------------
# Gradio Interface
# ----------------------------
iface = gr.Interface(
fn=detect_image,
inputs=gr.Image(type="pil"),
outputs=[gr.Textbox(label="Hasil Deteksi"), gr.Number(label="Confidence (%)")],
title="AI vs Foto Asli Detector",
description="Unggah gambar, sistem akan mendeteksi apakah gambar asli atau dihasilkan AI menggunakan ensemble 3 model + voting."
)
iface.launch()
|