KhadijaAsehnoune12
commited on
Commit
•
54ca5bc
1
Parent(s):
4755ddc
Update app.py
Browse files
app.py
CHANGED
@@ -2,6 +2,7 @@ import gradio as gr
|
|
2 |
from transformers import pipeline
|
3 |
from huggingface_hub import hf_hub_download
|
4 |
import json
|
|
|
5 |
|
6 |
# Define model repository details
|
7 |
model_repo = "KhadijaAsehnoune12/LeafDiseaseDetector"
|
@@ -18,30 +19,9 @@ with open(config_path, "r", encoding="utf-8") as f:
|
|
18 |
pipe = pipeline(task="image-classification", model=model_repo)
|
19 |
|
20 |
# Define a custom prediction function
|
21 |
-
import numpy as np
|
22 |
-
|
23 |
-
from PIL import Image
|
24 |
-
from io import BytesIO
|
25 |
-
import numpy as np
|
26 |
-
import base64
|
27 |
-
|
28 |
def predict(image):
|
29 |
-
# Check if the input image is a base64 encoded string
|
30 |
-
if isinstance(image, str):
|
31 |
-
# Decode the base64 encoded image string and convert it to a PIL image object
|
32 |
-
image_data = BytesIO(base64.b64decode(image))
|
33 |
-
pil_image = Image.open(image_data)
|
34 |
-
# Convert the PIL image to a numpy array
|
35 |
-
image_np = np.array(pil_image)
|
36 |
-
elif isinstance(image, np.ndarray):
|
37 |
-
# If the input image is already a numpy array, use it directly
|
38 |
-
image_np = image
|
39 |
-
else:
|
40 |
-
# If the input is neither a base64 encoded string nor a numpy array, raise an error
|
41 |
-
raise ValueError("Input image must be either a base64 encoded string or a numpy array.")
|
42 |
-
|
43 |
# Get the predictions from the pipeline
|
44 |
-
predictions = pipe(
|
45 |
# Get the predicted label index
|
46 |
predicted_index = predictions[0]['label']
|
47 |
# Map the index to the corresponding disease name using id2label
|
@@ -50,15 +30,13 @@ def predict(image):
|
|
50 |
confidence_score = predictions[0]['score']
|
51 |
return f"{label_name} ({confidence_score:.2f})"
|
52 |
|
53 |
-
|
54 |
-
|
55 |
# Create Gradio interface
|
56 |
iface = gr.Interface(fn=predict,
|
57 |
-
inputs=gr.Image(type="numpy"),
|
58 |
-
outputs=gr.Textbox(),
|
59 |
title="Orange Disease Image Classification",
|
60 |
description="Detect diseases in orange leaves and fruits.",
|
61 |
-
examples=[
|
62 |
|
63 |
# Launch the app
|
64 |
iface.launch()
|
|
|
2 |
from transformers import pipeline
|
3 |
from huggingface_hub import hf_hub_download
|
4 |
import json
|
5 |
+
import numpy as np
|
6 |
|
7 |
# Define model repository details
|
8 |
model_repo = "KhadijaAsehnoune12/LeafDiseaseDetector"
|
|
|
19 |
pipe = pipeline(task="image-classification", model=model_repo)
|
20 |
|
21 |
# Define a custom prediction function
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
def predict(image):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
# Get the predictions from the pipeline
|
24 |
+
predictions = pipe(image)
|
25 |
# Get the predicted label index
|
26 |
predicted_index = predictions[0]['label']
|
27 |
# Map the index to the corresponding disease name using id2label
|
|
|
30 |
confidence_score = predictions[0]['score']
|
31 |
return f"{label_name} ({confidence_score:.2f})"
|
32 |
|
|
|
|
|
33 |
# Create Gradio interface
|
34 |
iface = gr.Interface(fn=predict,
|
35 |
+
inputs=gr.inputs.Image(type="numpy"),
|
36 |
+
outputs=gr.outputs.Textbox(),
|
37 |
title="Orange Disease Image Classification",
|
38 |
description="Detect diseases in orange leaves and fruits.",
|
39 |
+
examples=[np.random.rand(224, 224, 3)])
|
40 |
|
41 |
# Launch the app
|
42 |
iface.launch()
|