Update app.py

app.py

@@ -5,15 +5,16 @@ import numpy as np
 import tensorflow as tf
 import streamlit as st 
 import tempfile
-model = load_model('HandSignClassifier (2).h5')
+model = load_model('HandSignClassifier.h5')
 # Open the video file
 f = st.file_uploader("Choose a Video")
 array = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','del','nothing','space']
     # Read the video file from the file-like object
 if f is not None:
     img = Image.open(f)
-    img = img.resize((256,256))
-    img = np.reshape(img,(1,256,256,3))
+    img = img.resize((28,28))
+    img = img.convert('L')
+    img = np.reshape(img,(1,28,28,1))
     pred = model.predict(img)
     st.image(img,use_column_width=True)
     st.write(array[np.argmax(pred)])