Zeyadd-Mostaffa's picture
Upload 2 files
cba8635 verified
raw
history blame
3.39 kB
import os
import cv2
import numpy as np
import gradio as gr
import tensorflow as tf
from mtcnn import MTCNN
from huggingface_hub import hf_hub_download
from tensorflow.keras.models import load_model
from tensorflow.keras.applications.xception import preprocess_input as xcp_pre
from tensorflow.keras.applications.efficientnet import preprocess_input as eff_pre
# ---------------------------------------------------------
# Load models from Hugging Face Hub
# ---------------------------------------------------------
xcp_path = hf_hub_download(repo_id="Zeyadd-Mostaffa/deepfake-image-detector_final", filename="xception_model.h5")
eff_path = hf_hub_download(repo_id="Zeyadd-Mostaffa/deepfake-image-detector_final", filename="efficientnet_model.h5")
xcp_model = load_model(xcp_path)
eff_model = load_model(eff_path)
# ---------------------------------------------------------
# Face Detection
# ---------------------------------------------------------
detector = MTCNN()
def extract_faces(image):
faces = detector.detect_faces(image)
if not faces:
return []
results = []
for i, face in enumerate(faces):
x, y, w, h = face['box']
x, y = max(0, x), max(0, y)
cropped = image[y:y+h, x:x+w]
if cropped.shape[0] >= 60 and cropped.shape[1] >= 60:
results.append((cropped, (x, y, w, h)))
return results
# ---------------------------------------------------------
# Inference Function
# ---------------------------------------------------------
def predict_faces(image):
faces = extract_faces(image)
if not faces:
return "No faces detected", None
annotated = image.copy()
results = []
for i, (face, (x, y, w, h)) in enumerate(faces):
# Preprocess
xcp_img = cv2.resize(face, (299, 299))
eff_img = cv2.resize(face, (224, 224))
xcp_tensor = xcp_pre(xcp_img.astype(np.float32))[np.newaxis, ...]
eff_tensor = eff_pre(eff_img.astype(np.float32))[np.newaxis, ...]
# Predict
xcp_pred = xcp_model.predict(xcp_tensor, verbose=0).flatten()[0]
eff_pred = eff_model.predict(eff_tensor, verbose=0).flatten()[0]
avg_pred = (xcp_pred + eff_pred) / 2
label = "Real" if avg_pred > 0.5 else "Fake"
confidence = f"{avg_pred:.2f}"
# Annotate
color = (0, 255, 0) if label == "Real" else (0, 0, 255)
cv2.rectangle(annotated, (x, y), (x + w, y + h), color, 2)
cv2.putText(annotated, f"{label} ({confidence})", (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, color, 2)
results.append(f"Face {i+1}: {label} (Avg: {avg_pred:.3f}, XCP: {xcp_pred:.3f}, EFF: {eff_pred:.3f})")
return "\n".join(results), annotated
# ---------------------------------------------------------
# Gradio Interface
# ---------------------------------------------------------
interface = gr.Interface(
fn=predict_faces,
inputs=gr.Image(type="numpy", label="Upload Image"),
outputs=[
gr.Textbox(label="Predictions"),
gr.Image(type="numpy", label="Annotated Image")
],
title="Deepfake Detector (Multi-Face Ensemble)",
description="This model detects all faces in an image and classifies each one as real or fake using Xception and EfficientNetB4 ensemble."
)
interface.launch()