GUI fix

Experiments/Baseline/GUI.py

@@ -4,6 +4,7 @@ import json
 from PIL import Image
 from google.oauth2 import service_account
 from baseline_utils import detect_text_in_image, summarize_diary_text, analyze_writer_image, generate_comic_book
+import glob
 
 # Load secrets
 openai_api_key = st.secrets["general"]["openai_api_key"]
@@ -24,13 +25,16 @@ uploaded_writer_image = st.file_uploader("Upload a photo of the writer", type=["
 if uploaded_diary and uploaded_writer_image:
     st.write("Analyzing your diary and writer...")
 
-    # Read the uploaded images
+    # Read the uploaded images using file-like objects
     diary_image = Image.open(uploaded_diary)
     writer_image = Image.open(uploaded_writer_image)
 
-    # Save uploaded images temporarily (Streamlit does this automatically with file objects)
-    diary_image_path = uploaded_diary.name
-    writer_image_path = uploaded_writer_image.name
+    # Save the file-like objects as image files (optional if needed)
+    diary_image_path = "temp_diary_image.png"
+    writer_image_path = "temp_writer_image.png"
+
+    diary_image.save(diary_image_path)
+    writer_image.save(writer_image_path)
 
     # Detect text from the diary image
     google_credentials = get_google_credentials()
@@ -46,4 +50,11 @@ if uploaded_diary and uploaded_writer_image:
     st.write("Generating comic book images...")
     generate_comic_book(summarized_text, writer_summary, num_pages=5)
 
-    st.write("Comic book generated successfully!")
+    st.write("Comic book generated successfully!")
+
+    # Assuming generated images are saved as 'comic_book/page_1.png', 'comic_book/page_2.png', etc.
+    image_files = sorted(glob.glob("comic_book/page_*.png"))  # Find all the generated comic book pages
+
+    for image_file in image_files:
+        # Display each comic book page
+        st.image(image_file, caption=image_file.split('/')[-1], use_column_width=True)

Experiments/Baseline/baseline_utils.py

@@ -9,9 +9,7 @@ import os
 
 # Utilize the Google Cloud Vision API to recognize text in the
 # input input_images (diary input_images), https://cloud.google.com/vision.
-def detect_text_in_image(image_path, credentials_path):
-    # Load the service account key from the credentials JSON file
-    credentials = service_account.Credentials.from_service_account_file(credentials_path)
+def detect_text_in_image(image_path, credentials):
 
     # Create a Vision API client using the credentials
     client = vision.ImageAnnotatorClient(credentials=credentials)

requirements.txt

@@ -3,4 +3,5 @@ google-cloud-vision
 google-auth
 google-generativeai
 diffusers
-torch
+torch
+streamlit