Update app.py

app.py

@@ -1,45 +1,26 @@
 # app.py
 
-import gradio as gr
-import numpy as np
-import tensorflow_hub as hub
-from tensorflow.keras.models import load_model
-import cv2
-import os
-
-# Disable GPU (optional, for CPU-only execution)
-os.environ["CUDA_VISIBLE_DEVICES"] = "-1"
+__all__ = ['bird', 'learn', 'classify_image', 'categories', 'image', 'label', 'examples', 'intf']
 
-# Define a dictionary to map the custom layer to its implementation
-custom_objects = {'KerasLayer': hub.KerasLayer}
+# Load categories from label.txt
+with open('label.txt', 'r') as file:
+    categories = [line.strip() for line in file.readlines()]
 
-# Load your model (ensure the path is correct)
-try:
-    model = load_model('bird_model.h5', custom_objects=custom_objects)
-except Exception as e:
-    print(f"Model loading failed with error: {e}")
-    exit(1)
-
-# Define your class labels or categories for predictions
-train_info = []  # Replace with your actual class labels
+# Load the model
+from fastai.vision.all import *
+import gradio as gr
 
-# Read image names from the text file
-with open('labelwithspace.txt', 'r') as file:
-    train_info = [line.strip() for line in file.read().splitlines()]
+learn = load_learner('model.pkl')
 
-# Preprocess and predict function
-def predict_image(image):
-    img = cv2.resize(image, (224, 224))  # Resize to match model input size
-    img = img / 255.0  # Normalize pixel values
-    img = img[np.newaxis, ...]  # Add batch dimension
-    predictions = model.predict(img)[0]  # Get predictions
-    top_class = np.argmax(predictions)  # Get the index of the top prediction
-    label = train_info[top_class]  # Use the index to retrieve the label
-    return label
+# Define the classification function
+def classify_image(img):
+    preds, idx, probs = learn.predict(img)
+    return dict(zip(categories, map(float, probs)))
 
-# Define Gradio interface
-input_image = gr.Image(shape=(224, 224), label="Input Image")
-output_label = gr.Label(label="Predicted Bird Species")
+# Gradio components
+image = gr.components.Image()
+label = gr.components.Label(num_top_classes=3)
 
-# Launch Gradio
-gr.Interface(fn=predict_image, inputs=input_image, outputs=output_label).launch()
+# Create and launch the interface
+intf = gr.Interface(fn=classify_image, inputs=image, outputs=label)
+intf.launch(inline=False, share=True)