app.py

# import part
import streamlit as st
from transformers import pipeline

# function part
# image2text
def img2text(img):
    image_to_text_model = pipeline("image-to-text",
                                   model="nlpconnect/vit-gpt2-image-captioning")
    text = image_to_text_model(img)[0]["generated_text"]
    return text

# text2story
def text2story(text):
    text_generation_model = pipeline("text-generation",
                                   model="openai-community/gpt2")
    story_text = f"Once upon a time in a land far, far away, {text}"
    generated_story = text_generation_model(story_text,
                            max_length=100,
                            num_return_sequences=1)
    return generated_story[0]['generated_text']
        
# text2audio
def text2audio(story_text):
    text_to_speech_model = pipeline("text-to-speech", model="facebook/mms-tts-eng")
    speech_output = text_to_speech_model(story_text)
    return speech_output

# main part
st.set_page_config(page_title="Your Image to Audio Story",
                    page_icon="*")
st.header("Turn Your Image to Audio Story")
uploaded_file = st.file_uploader("Select an Image...", type=["jpg", "png", "jpeg"])

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_container_width=True)

    # stage 1
    st.text('Processing img2text...')
    scenario = img2text(uploaded_file.name)
    st.write(scenario)

     # stage 2
    st.text('Generating a story...')
    generated_story = text2story(scenario)  
    # Use the scenario from img2text
    st.write(generated_story)

     # stage 3
    st.text('Generating audio data...')
    audio_data = text2audio(generated_story)

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