app.py

import gradio as gr
from fastai.vision.all import *
from PIL import Image
import io
import numpy as np


learn = load_learner('model.pkl')
labels = learn.dls.vocab
input_shape = (512, 512)  # Dimensões desejadas para a entrada

def preprocess_image(img):
    img = Image.fromarray(np.uint8(img))
    img = img.resize(input_shape)
    return img

def predict(img):
    img = preprocess_image(img)
    img = PILImage.create(img)
    pred, pred_idx, probs = learn.predict(img)
    return {labels[i]: float(probs[i]) for i in range(len(labels))}

title = "Tuberculosis detector"
description = "A tuberculosis detector trained on 3700 chest X-ray images. Created as a demo for Gradio and HuggingFace Spaces."
examples = ['1.jpeg', '2.jpeg', '3.jpeg', '4.jpeg']
interpretation = 'default'
enable_queue = True

gr.Interface(fn=predict, inputs=gr.inputs.Image(), outputs=gr.outputs.Label(num_top_classes=2),
             title=title, description=description, interpretation=interpretation,
             examples=examples, enable_queue=enable_queue).launch()