File size: 1,322 Bytes
92aa748
 
a2a5b20
92aa748
26573f5
cf5eea4
92aa748
 
 
 
 
 
 
 
d564473
 
 
 
 
26e91fb
2e04135
 
 
944da42
658b9fc
4a1ed7b
92aa748
 
 
 
26e91fb
187b87b
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
import gradio as gr
from fastai.vision.all import *
from efficientnet_pytorch import EfficientNet 

#learn = load_learner('model/predictcovidfastaifinal18102023.pkl')
learn = load_learner('model/final_20102023_eb7_model.pkl')

categories = learn.dls.vocab

def predict_image(get_image):
   pred, idx, probs = learn.predict(get_image)
   return dict(zip(categories, map(float, probs)))

title = "Detect COVID_19 Infection Xray Chest Images"
description = """
This Space demonstrates model based on efficientnet base model. 

The model is trained using [anasmohammedtahir/covidqu] (https://www.kaggle.com/datasets/anasmohammedtahir/covidqu)dataset
"""
examples = [[
    ['covid/covid_1038.png'], ['covid/covid_1034.png'], 
    ['covid/cd.png'], ['covid/covid_1021.png'], 
    ['covid/covid_1027.png'], ['covid/covid_1042.png'], 
    ['covid/covid_1031.png']]
  
article="<p style='text-align: center'><a href='https://www.kaggle.com/datasets/anasmohammedtahir/covidqu' target='_blank'>COVID-QU-Ex Dataset</a></p>"
interpretation='default'
enable_queue=True

gr.Interface(fn=predict_image, inputs=gr.Image(shape=(224,224)),
             outputs = gr.Label(num_top_classes=3),title=title,description=description,examples=examples, article=article, interpretation=interpretation,enable_queue=enable_queue).launch(share=False)