Spaces:
Build error
Build error
File size: 1,267 Bytes
9c91b26 667c6df 9c91b26 d129e85 667c6df f3a9b48 36ae1e2 9c91b26 36ae1e2 9c91b26 36ae1e2 9c91b26 f3a9b48 9c91b26 cad188d cce1cb0 36ae1e2 9c91b26 f6e90f5 c7f3283 | 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 | from fastai.vision.all import *
import gradio as gr
from PIL import Image, ImageChops, ImageEnhance
import cv2
#Converting the images to ELA format for prediction
def convert_to_ela_image(img_array, quality):
resaved_filename = 'temp.jpg'
img = Image.fromarray(img_array)
im = img.convert('RGB')
im.save(resaved_filename, 'JPEG', quality=quality)
resaved_im = Image.open(resaved_filename)
ela_im = ImageChops.difference(im, resaved_im)
extrema = ela_im.getextrema()
max_diff = max([ex[1] for ex in extrema])
if max_diff == 0:
max_diff = 1
scale = 255.0 / max_diff
ela_im = ImageEnhance.Brightness(ela_im).enhance(scale)
ela_im = ela_im.resize((256, 256))
return ela_im
#Importing the model
learn = load_learner('resnet34_ela_images.pkl')
categories = ('Fake', 'Real')
#Defining the function to get the prediction values
def classify_image(img):
im = convert_to_ela_image(img, 90)
pred,idx,probs = learn.predict(im)
return dict(zip(categories, map(float,probs)))
image = gr.Image()
label = gr.Label()
examples = ['lion_f.png', 'boat_r.jpg', 'potato_f.jpg']
intf = gr.Interface(fn=classify_image, inputs=image, outputs = label, examples = examples)
intf.launch(inline=False) |