| |
| import streamlit as st |
| from transformers import pipeline |
| from PIL import Image |
| import io |
| from scipy.io.wavfile import write as write_wav |
|
|
| def generate_image_caption(image): |
| """Generates a caption for the given image""" |
| img2caption = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base") |
| result = img2caption(image) |
| return result[0]['generated_text'] |
|
|
| def text2story(text): |
| """Generates a children's story from text input""" |
| story_prompt = f"Create a funny 100-word story for 8-year-olds about: {text}. Include: 1) A silly character 2) Magical object 3) Sound effects 4) Happy ending" |
| |
| pipe = pipeline("text-generation", model="pranavpsv/genre-story-generator-v2") |
| story_text = pipe( |
| story_prompt, |
| max_new_tokens=200, |
| temperature=0.9, |
| top_k=50 |
| )[0]['generated_text'] |
| return story_text.split("Happy ending")[-1].strip() |
|
|
| def story_to_speech(story_text): |
| """Converts story text to audio using TTS""" |
| tts_pipe = pipeline("text-to-speech", model="suno/bark-small") |
| audio_output = tts_pipe(story_text[:400]) |
| |
| |
| audio_bytes = io.BytesIO() |
| audio_np = (audio_output["audio"] * 32767).astype(np.int16) |
| write_wav(audio_bytes, audio_output["sampling_rate"], audio_np) |
| audio_bytes.seek(0) |
| |
| return audio_bytes |
|
|
| def main(): |
| st.title("📖 Image Story Generator with Audio") |
| st.write("Upload an image to get a magical story read aloud!") |
|
|
| uploaded_image = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"]) |
|
|
| if uploaded_image: |
| image = Image.open(uploaded_image).convert("RGB") |
| st.image(image, use_column_width=True) |
| |
| with st.spinner("✨ Analyzing image..."): |
| caption = generate_image_caption(image) |
| |
| st.subheader("Image Understanding") |
| st.write(caption) |
| |
| with st.spinner("📖 Writing story..."): |
| story = text2story(caption) |
| |
| st.subheader("Magical Story") |
| st.write(story) |
| |
| if st.button("🎧 Read Story Aloud"): |
| with st.spinner("🔊 Generating audio..."): |
| try: |
| audio_bytes = story_to_speech(story) |
| st.audio(audio_bytes, format="audio/wav") |
| except Exception as e: |
| st.error(f"Error generating audio: {str(e)}") |
|
|
| if __name__ == "__main__": |
| main() |
