Create app.py

app.py

@@ -0,0 +1,37 @@
+import streamlit as st
+from PIL import Image
+import numpy as np
+
+def mock_encoder(image):
+    """Simulates encoding an image into a latent representation."""
+    return np.random.normal(0, 1, (1, 2)), np.random.normal(0, 1, (1, 2)), np.random.normal(0, 1, (1, 2))
+
+def mock_decoder(latent_representation):
+    """Simulates decoding a latent representation back into an image."""
+    return np.random.rand(28, 28, 1) * 255  # Random image for demonstration
+
+def latent_space_augmentation(image, encoder, decoder, noise_scale=0.1):
+    z_mean, z_log_var, _ = encoder(image)
+    epsilon = np.random.normal(size=z_mean.shape)
+    z_augmented = z_mean + np.exp(0.5 * z_log_var) * epsilon * noise_scale
+    augmented_image = decoder(z_augmented)
+    return np.squeeze(augmented_image)
+
+st.title("VAE-based Image Augmentation Demo")
+
+uploaded_image = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])
+if uploaded_image is not None:
+    image = Image.open(uploaded_image).convert('L')
+    image = image.resize((28, 28))  # Resizing for simplicity
+    st.image(image, caption="Uploaded Image", use_column_width=True)
+    
+    if st.button("Augment Image"):
+        # Convert PIL image to numpy array
+        image_array = np.array(image) / 255.0  # Normalize the image
+        image_array = image_array.reshape((28, 28, 1))  # Reshape for the mock encoder/decoder
+        
+        # Perform augmentation
+        augmented_image = latent_space_augmentation(image_array, mock_encoder, mock_decoder)
+        
+        # Display the augmented image
+        st.image(augmented_image, caption="Augmented Image", clamp=True, use_column_width=True)