Spaces:
Sleeping
Sleeping
Upload 3 files
Browse files- CNN_Prak4_ML.h5 +3 -0
- app.py +63 -0
- requirement.txt +3 -0
CNN_Prak4_ML.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fd8b10e96ef8fa2e2d16c3251e2b21d6d9ea0942dcaea4941ee632dcce219f4c
|
3 |
+
size 1781096
|
app.py
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import tensorflow as tf
|
3 |
+
from tensorflow.keras.preprocessing.image import load_img, img_to_array
|
4 |
+
import numpy as np
|
5 |
+
from PIL import Image
|
6 |
+
import io
|
7 |
+
|
8 |
+
st.set_page_config(
|
9 |
+
page_title="Waste Classifier",
|
10 |
+
layout="centered"
|
11 |
+
)
|
12 |
+
|
13 |
+
@st.cache_resource
|
14 |
+
def load_model():
|
15 |
+
return tf.keras.models.load_model('CNN_Prak4_ML.h5')
|
16 |
+
|
17 |
+
def preprocess_image(img):
|
18 |
+
img = img.resize((244, 244))
|
19 |
+
img = img_to_array(img)
|
20 |
+
img = np.expand_dims(img, axis=0)
|
21 |
+
img = img / 255.0
|
22 |
+
return img
|
23 |
+
|
24 |
+
LABEL_CLASS = {
|
25 |
+
0: "Cardboard",
|
26 |
+
1: "Glass",
|
27 |
+
2: "Metal",
|
28 |
+
3: "Paper",
|
29 |
+
4: "Textile Trash",
|
30 |
+
5: "Vegetation"
|
31 |
+
}
|
32 |
+
|
33 |
+
|
34 |
+
def main():
|
35 |
+
st.title("Waste Classifier")
|
36 |
+
st.write("Upload an image and the model will predict waste image")
|
37 |
+
|
38 |
+
uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])
|
39 |
+
|
40 |
+
if uploaded_file is not None:
|
41 |
+
image = Image.open(uploaded_file)
|
42 |
+
st.image(image, caption='Uploaded Image', use_column_width=True)
|
43 |
+
|
44 |
+
if st.button('Predict'):
|
45 |
+
model = load_model()
|
46 |
+
|
47 |
+
processed_image = preprocess_image(image)
|
48 |
+
|
49 |
+
with st.spinner('Predicting...'):
|
50 |
+
prediction = model.predict(processed_image)
|
51 |
+
pred_class = LABEL_CLASS[np.argmax(prediction)]
|
52 |
+
confidence = float(prediction.max()) * 100
|
53 |
+
|
54 |
+
st.success(f'Prediction: {pred_class.upper()}')
|
55 |
+
st.info(f'Confidence: {confidence:.2f}%')
|
56 |
+
|
57 |
+
st.write("Class Probabilities:")
|
58 |
+
for i, prob in enumerate(prediction[0]):
|
59 |
+
st.progress(float(prob))
|
60 |
+
st.write(f"{LABEL_CLASS[i]}: {float(prob)*100:.2f}%")
|
61 |
+
|
62 |
+
if __name__ == "__main__":
|
63 |
+
main()
|
requirement.txt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
pillow==10.4.0
|
2 |
+
streamlit==1.39.0
|
3 |
+
tensorflow==2.18.0
|