Upload 3 files

dog_breed.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5978d94fb5b9380b71e8c60f28b62187f1e2750048c73bdf2ef465287e90c47d
+size 2022552

main_app.py

@@ -0,0 +1,42 @@
+#Library imports
+import numpy as np
+import streamlit as st
+import cv2
+from keras.models import load_model
+
+
+#Loading the Model
+model = load_model('dog_breed.h5')
+
+#Name of Classes
+CLASS_NAMES = ['Scottish Deerhound','Maltese Dog','Bernese Mountain Dog']
+
+#Setting Title of App
+st.title("Dog Breed Prediction")
+st.markdown("Upload an image of the dog")
+
+#Uploading the dog image
+dog_image = st.file_uploader("Choose an image...", type="png")
+submit = st.button('Predict')
+#On predict button click
+if submit:
+
+
+    if dog_image is not None:
+
+        # Convert the file to an opencv image.
+        file_bytes = np.asarray(bytearray(dog_image.read()), dtype=np.uint8)
+        opencv_image = cv2.imdecode(file_bytes, 1)
+
+
+
+        # Displaying the image
+        st.image(opencv_image, channels="BGR")
+        #Resizing the image
+        opencv_image = cv2.resize(opencv_image, (224,224))
+        #Convert image to 4 Dimension
+        opencv_image.shape = (1,224,224,3)
+        #Make Prediction
+        Y_pred = model.predict(opencv_image)
+
+        st.title(str("The Dog Breed is "+CLASS_NAMES[np.argmax(Y_pred)]))

requirements.txt

@@ -0,0 +1,4 @@
+Keras==2.4.3
+opencv_python==4.4.0.46
+numpy==1.18.5
+streamlit==0.71.0