Spaces:
Sleeping
Sleeping
| # import part | |
| import streamlit as st | |
| from transformers import pipeline | |
| from gtts import gTTS # Using gTTS for text-to-speech | |
| import os | |
| import re # For removing unwanted words | |
| # function part | |
| # 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"] | |
| # Make the caption more fun and descriptive | |
| fun_caption = f"Wow! {text.capitalize()}! 🌟" | |
| return fun_caption | |
| # text2story | |
| def text2story(text): | |
| # Use facebook/bart-large-cnn for text generation | |
| story_generator = pipeline("text-generation", model="facebook/bart-large-cnn") | |
| # Add a more explicit prompt to guide the story generation | |
| prompt = f"Write a fun and realistic story for kids based on this: {text}. The story should be under 95 words and suitable for children aged 3-10. Avoid using the word 'illustration'." | |
| story = story_generator(prompt, max_length=95, num_return_sequences=1)[0]["generated_text"] | |
| # Remove the prompt from the generated story | |
| story = story.replace(prompt, "").strip() | |
| # Remove the word "illustration" (case-insensitive) using regex | |
| story = re.sub(r"\billustration\b", "", story, flags=re.IGNORECASE) | |
| return story[:95] # Limit to 95 words | |
| # text2audio | |
| def text2audio(story_text): | |
| # Use gTTS for text-to-speech conversion | |
| tts = gTTS(text=story_text, lang="en") | |
| audio_file = "story_audio.mp3" | |
| tts.save(audio_file) | |
| return audio_file | |
| # main part | |
| st.set_page_config(page_title="Story Explorer", page_icon="🦜") | |
| st.header("Story Explorer: Turn Your Picture into a Fun Story! 🎨📖") | |
| uploaded_file = st.file_uploader("Choose a picture...", type=["jpg", "png", "jpeg"]) | |
| if uploaded_file is not None: | |
| bytes_data = uploaded_file.getvalue() | |
| with open(uploaded_file.name, "wb") as file: | |
| file.write(bytes_data) | |
| st.image(uploaded_file, caption="Your Picture", use_container_width=True) | |
| # Stage 1: Image to Text | |
| st.text('Let’s explore your picture! 🧐✨') | |
| scenario = img2text(uploaded_file.name) | |
| st.write(f"**Here’s what I see in your picture:** {scenario}") | |
| # Stage 2: Text to Story | |
| st.text('Creating a fun story for you! 📖') | |
| story = text2story(scenario) | |
| st.write(f"**Here’s your story:** {story}") | |
| # Stage 3: Story to Audio data | |
| st.text('Turning your story into audio... 🎧') | |
| audio_file = text2audio(story) | |
| # Play button | |
| if st.button("Play Audio"): | |
| st.audio(audio_file, format="audio/mp3") | |
| # Clean up the audio file after use | |
| os.remove(audio_file) |