Update app.py

app.py

@@ -1,4 +1,6 @@
 import streamlit as st
+import time
+
 import cv2
 import pandas
 from PIL import Image
@@ -13,31 +15,37 @@ st.markdown("This is a Deep Learning application to identify if a satellite imag
 st.markdown('The predicting result will be "Jacaranda", or "Others".')
 st.markdown('You can click "Browse files" multiple times until adding all images before generating prediction.\n')
 
-
-uploaded_file = st.file_uploader("Upload image files", type="jpg", accept_multiple_files=True)
-
-#image_iterator = paginator("Select a page", uploaded_file)
-#indices_on_page, images_on_page = map(list, zip(*image_iterator))
-st.image(uploaded_file, width=100)
-
 img_height = 224
 img_width = 224
 class_names = ['Jacaranda', 'Others']
 
 model = tf.keras.models.load_model('model')
 
+state = st.session_state
+if 'file_uploader_key' not in state:
+    state['file_uploader_key'] = 0
+
+if "uploaded_files" not in state:
+    state["uploaded_files"] = []
+
+uploaded_file = st.file_uploader(
+    "Upload images", 
+    type="jpg" or 'jpeg' or 'bmp' or 'png' or 'tif', 
+    accept_multiple_files=True, 
+    key=state['file_uploader_key'])
+
 if uploaded_file is not None:
-    #n = len(uploaded_file)
-    #row_size = 5
-    #grid = st.columns(row_size)
-    #col = 0
-    Generate_pred = st.button("Generate Prediction")
-    if Generate_pred:
+    st.image(uploaded_file, width=100)
+    state["uploaded_files"] = uploaded_file
+
+    if st.button("Clear uploaded images"):
+        state["file_uploader_key"] += 1
+        st.empty()
+        time.sleep(.5)
+        st.experimental_rerun()
+    
+    if st.button("Generate prediction"):
         for file in uploaded_file:
-        #    with grid[col]:
-        #        img = Image.open(file)
-        #        st.image(img)
-        #col += 1
             img = Image.open(file)
             img_array = img_to_array(img)
             img_array = tf.expand_dims(img_array, axis = 0) # Create a batch
@@ -45,4 +53,5 @@ if uploaded_file is not None:
     
             predictions = model.predict(processed_image)
             score = predictions[0]
-            st.markdown("Predicted class of the image {} is : {}".format(file, class_names[np.argmax(score)]))
+            st.markdown("Predicted class of the image {} is : {}".format(file, class_names[np.argmax(score)]))
+