app.py

from transformers import pipeline
import streamlit as st
from PIL import Image

# img2text
def img2text(url):
    image_to_text_model = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
    text = image_to_text_model(url)[0]["generated_text"]
    return text
    
# text2story
def text2story(text):
    text_to_story_model = pipeline("text-generation", model="distilbert/distilgpt2")
    if isinstance(text, list):
        text="".join(text)# Ensure input is a single string
    story_text = text_to_story_model(text, max_length=100, num_return_sequences=1)
    return story_text[0]['generated text']

# text2audio
def text2audio(story_text):
    text_to_audio_model = pipeline("text-to-speech", model="facebook/mms-tts-eng")
    audio_data = text_to_audio_model(story_text)
    return audio_data
    
#main part
st.set_page_config(page_title="Your Image to Audio Story",
                   page_icon="🦜")
st.header("Turn Your Image to Story")
uploaded_file= st.file_uploader("Select an Image...")

if uploaded_file is not None:
    print(uploaded_file)
    bytes_data = uploaded_file.getvalue()
    with open(uploaded_file.name,"wb") as file:
        file.write(bytes_data)
    st.image(uploaded_file,caption="Uploaded Image",
             use_column_width=True)
    
#Stage 1:Image to Text
st.text('Processing img2text...')
scenario = img2text(uploaded_file)
st.write(scenario)

#Stage 2: Text to Story
st.text('Generating a story...')
story = text2story(scenario)
st.write(story)

#Stage 3:Story to Audio data
st.text('Generating audio data...')
audio_data =text2audio(story)

# Play button
if st.button("Play Audio"):
    st.audio(audio_data['audio'],
                format="audio/wav",
                start_time=0,
                sample_rate = audio_data['sampling_rate'])