Update app.py

app.py

@@ -2,10 +2,11 @@ import streamlit as st
 from PIL import Image
 import torch
 import easyocr
+import numpy as np
 import io
 from transformers import CLIPModel, CLIPImageProcessor
 
-# ✅ Fix: set_page_config() must be the first command
+# ✅ Fix: set_page_config() must be the first Streamlit command
 st.set_page_config(page_title="Multimodal AI Assistant", layout="wide")
 
 # ---- Load CLIP Model (Vision Only) ---- #
@@ -13,7 +14,7 @@ st.set_page_config(page_title="Multimodal AI Assistant", layout="wide")
 def load_clip_model():
     model = CLIPModel.from_pretrained(
         "fxmarty/clip-vision-model-tiny", 
-        ignore_mismatched_sizes=True  # Fix size mismatch
+        ignore_mismatched_sizes=True  # ✅ Fix size mismatch
     )
     processor = CLIPImageProcessor.from_pretrained("fxmarty/clip-vision-model-tiny")
     return model, processor
@@ -36,18 +37,17 @@ uploaded_file = st.file_uploader("📤 Upload an image", type=["jpg", "png", "jp
 
 if uploaded_file is not None:
     # Convert file to image format
-    image = Image.open(uploaded_file)
+    image = Image.open(uploaded_file).convert("RGB")
 
     # ✅ Fix: use `use_container_width` instead of `use_column_width`
     st.image(image, caption="Uploaded Image", use_container_width=True)
 
-    # Convert uploaded file to bytes for EasyOCR
-    file_bytes = uploaded_file.read()  # ✅ Fix: Convert file to bytes
-    image_bytes = io.BytesIO(file_bytes)
+    # ✅ Convert PIL image to NumPy array for EasyOCR
+    image_np = np.array(image)
 
-    # Extract Text using OCR
+    # ✅ Fix: Pass the correct format to EasyOCR
     with st.spinner("🔍 Extracting text from image..."):
-        extracted_text = reader.readtext(image_bytes, detail=0)
+        extracted_text = reader.readtext(image_np, detail=0)
 
     st.write("### 📝 Extracted Text:")
     if extracted_text: