Update app.py

app.py

@@ -1,29 +1,25 @@
 import gradio as gr
-from tensorflow.keras.preprocessing.image import ImageDataGenerator
 from tensorflow.keras.models import load_model
 import pandas as pd
 import numpy as np
-
-
-from tensorflow.keras.preprocessing import image
-
-def display_image_and_prediction(Oiseau):
+import cv2
+ 
+dataCsv = pd.read_csv('birds.csv')
+ 
+def display_image_and_prediction(bird):
     model = load_model('mon_modele/mon_modele')
     generator = np.load('test_data.npy')
-    dataCsv = pd.read_csv('birds.csv')
-
-    Oiseau_resized= np.resize(Oiseau,(224, 224))
-    #img_array = image.img_to_array(img)
-    #img_array = np.array(Oiseau_resized)
-    img_array = np.expand_dims(Oiseau_resized, axis=0)
-    img_array = img_array.astype('float32') / 255.0
-    
-    prediction = model.predict(img_array)
+ 
+    bird_array_resized = cv2.resize(bird,(224, 224), 3)
+    bird_array = np.expand_dims(bird_array_resized.astype('float32') / 255.0, axis=0)
+    prediction = model.predict(bird_array)
+     
     predicted_class = np.argmax(prediction)
     predicted_class_name = generator[predicted_class]
+ 
     scientificName = dataCsv.loc[dataCsv["labels"]==predicted_class_name,"scientific name"].iloc[0]#récupération du nom scientifique de cette oiseau
-
-    return predicted_class, scientificName
-
+ 
+    return predicted_class_name, scientificName
+ 
 iface = gr.Interface(fn=display_image_and_prediction, inputs="image", outputs=["text","text"])
-iface.launch()
+iface.launch(share=True)