| from PIL import Image |
| import requests |
| from io import BytesIO |
| from transformers import pipeline |
| import streamlit as st |
|
|
| def predict(image): |
| type_food = oracle(image, "What type of food is this?") |
| cal_est = oracle(image, "About how many calories are in this meal?") |
| guess1, guess2 = int(cal_est[0]['answer']), int(cal_est[1]['answer']) |
| return f"This is {type_food[0]['answer']}. I estimate this to contain {min(guess1, guess2)}-{max(guess1, guess2)} calories" |
| oracle = pipeline(model="dandelin/vilt-b32-finetuned-vqa") |
|
|
| def main(): |
| st.title("Image Question Answering App") |
| st.write("Upload an image and ask a question to get answers!") |
|
|
| |
| uploaded_image = st.file_uploader("Upload Image", type=["jpg", "jpeg", "png"]) |
| |
| if uploaded_image is not None: |
| image = Image.open(uploaded_image) |
| st.image(image, caption="Uploaded Image", use_column_width=True) |
| response = predict(image) |
| st.write(response) |
| |
| if __name__ == "__main__": |
| main() |
|
|
