Update app.py

app.py

@@ -5,17 +5,11 @@ import uuid
 from datetime import datetime
 import shutil
 from huggingface_hub import HfApi, create_repo, upload_file, upload_folder
-
-# Create directories for data storage
 os.makedirs("uploaded_images", exist_ok=True)
 os.makedirs("submissions", exist_ok=True)
-
-# Hugging Face Dataset configuration
 HF_TOKEN = os.environ.get("HF_TOKEN") 
 DATASET_NAME = "minemaster01/se-culture-dataset-results" 
 DATASET_CREATED = False
-
-# States by country dictionary
 states_by_country = {
     "India": [
         "Andhra Pradesh", "Arunachal Pradesh", "Assam", "Bihar", "Chhattisgarh", "Goa", "Gujarat", "Haryana",
@@ -72,32 +66,22 @@ def setup_hf_dataset():
             print(f"Error setting up dataset: {e}")
     elif not HF_TOKEN:
         print("Warning: HF_TOKEN not set. Data will be stored locally only.")
-
 def update_state_dropdown(country):
     """Update state dropdown based on selected country"""
     if country in states_by_country:
         return gr.Dropdown(choices=states_by_country[country], label=f"State/Province in {country}:", interactive=True)
     return gr.Dropdown(choices=[], label="State/Province:", interactive=True)
-
 def process_submission(input_img, language, country, state, city, se_asia_relevance, culture_knowledge, native_caption, english_caption,email):
-    # Generate unique ID for this submission
     submission_id = str(uuid.uuid4())
     timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
-    
-    # Save the image if provided
     image_path = None
     if input_img is not None:
-        # Create filename with submission ID
         image_filename = f"{timestamp}.jpg"
         image_path = os.path.join("uploaded_images", image_filename)
-        
-        # Save the image
-        if isinstance(input_img, str):  # If it's a file path
+        if isinstance(input_img, str):
             shutil.copy(input_img, image_path)
-        else:  # If it's a PIL Image
+        else:
             input_img.save(image_path)
-    
-    # Create a data structure for the submission
     submission_data = {
         "id": submission_id,
         "timestamp": timestamp,
@@ -112,18 +96,13 @@ def process_submission(input_img, language, country, state, city, se_asia_releva
         "english_caption": english_caption,
         "email": email
     }
-    
-    # Save the data as JSON
     json_filename = f"{timestamp}.json"
     json_path = os.path.join("submissions", json_filename)
     with open(json_path, "w") as f:
         json.dump(submission_data, f, indent=2)
-    
-    # Upload to Hugging Face Dataset if token is available
     if HF_TOKEN and DATASET_CREATED:
         try:
             api = HfApi()
-            # Upload the JSON data
             api.upload_file(
                 path_or_fileobj=json_path,
                 path_in_repo=f"submissions/{json_filename}",
@@ -131,7 +110,6 @@ def process_submission(input_img, language, country, state, city, se_asia_releva
                 repo_type="dataset",
                 token=HF_TOKEN
             )
-            # Upload the image if it exists
             if image_path and os.path.exists(image_path):
                 api.upload_file(
                     path_or_fileobj=image_path,
@@ -143,24 +121,15 @@ def process_submission(input_img, language, country, state, city, se_asia_releva
             print(f"Submission {submission_id} uploaded to Hugging Face Dataset")
         except Exception as e:
             print(f"Error uploading to dataset: {e}")
-    
-    # Return values to display in the interface
     location_info = f"Location: {city}, {state}, {country}" if state else f"Location: {city}, {country}"
     return input_img, f"Your text response: {language}", f"Selected location: {location_info}", f"SE Asia relevance: {se_asia_relevance}", f"Cultural knowledge source: {culture_knowledge}", f"Native caption: {native_caption}", f"English caption: {english_caption}"
-
 def clear_inputs():
     return None, "", None, None, "", None, None, "", "", ""
-
-# Initialize the dataset
 setup_hf_dataset()
-
 with gr.Blocks() as gradio_app:
     gr.Markdown("# South Asian Image Data Collection")
     gr.Markdown("Upload an image and answer questions about its cultural significance.")
-    
-    # Split the interface into two columns
     with gr.Row():
-        # Left column for image upload and basic information
         with gr.Column(scale=1):
             input_img = gr.Image(label="Upload an image", sources=['upload', 'webcam'], type="pil")
             language = gr.Dropdown(
@@ -169,8 +138,6 @@ with gr.Blocks() as gradio_app:
                 info="Select the native language relevant to the image",
                 interactive=True
             )
-            
-            # Location information in the left column
             country_dropdown = gr.Dropdown(
                 choices=["None","India", "Pakistan", "Bangladesh", "Afghanistan", "Bhutan", "Nepal", "Sri Lanka"],
                 label="Country where the image was taken:",
@@ -183,8 +150,6 @@ with gr.Blocks() as gradio_app:
             )
             city_textbox = gr.Textbox(label="City where the image was taken:", placeholder="Enter city name")
             email_input = gr.Textbox(label="Your Email:", placeholder="Enter your email address", info="Used as unique contributor ID")
-        
-        # Right column for additional information
         with gr.Column(scale=1):
             se_asia_relevance = gr.Radio(
                 choices=[
@@ -196,7 +161,6 @@ with gr.Blocks() as gradio_app:
                 ],
                 label="Is the image culturally relevant in South Asia?"
             )
-            
             culture_knowledge = gr.Radio(
                 choices=[
                     "I'm from this country/culture",
@@ -205,35 +169,26 @@ with gr.Blocks() as gradio_app:
                 label="How do you know about this culture?",
                 info="Please do not consult LLMs (e.g., GPT-4o, Claude, Command-R, etc.)"
             )
-            
             native_caption = gr.Textbox(label="Caption in Native Language:", placeholder="Enter caption in the native language of the culture depicted")
             english_caption = gr.Textbox(label="English Caption:", placeholder="Enter caption in English")
-    
-    # Buttons row
     with gr.Row():
         clear_btn = gr.Button("Clear")
         submit_btn = gr.Button("Submit")
-    
-    # Output display section - also split into two columns
     with gr.Row():
         with gr.Column(scale=1):
             output_img = gr.Image(label="Submitted Image")
             output_text = gr.Text(label="Text Response")
             output_location = gr.Text(label="Location Information")
-        
         with gr.Column(scale=1):
             output_relevance = gr.Text(label="South Asia Cultural Relevance")
             output_knowledge = gr.Text(label="Cultural Knowledge Source")
             output_native = gr.Text(label="Native Language Caption")
             output_english = gr.Text(label="English Caption")
-    
-    # Set up event handlers
     country_dropdown.change(
         fn=update_state_dropdown,
         inputs=country_dropdown,
         outputs=state_dropdown
     )
-    
     submit_btn.click(
         fn=process_submission,
         inputs=[
@@ -258,7 +213,6 @@ with gr.Blocks() as gradio_app:
             output_english
         ]
     )
-    
     clear_btn.click(
         fn=clear_inputs,
         inputs=[],
@@ -275,7 +229,5 @@ with gr.Blocks() as gradio_app:
             email_input 
         ]
     )
-
-
 if __name__ == "__main__":
     gradio_app.launch()