Spaces:
Sleeping
Sleeping
File size: 1,828 Bytes
034b2f2 2f5c91e 034b2f2 2f5c91e 034b2f2 2f5c91e 034b2f2 fee2e0a 2f5c91e 034b2f2 fee2e0a 034b2f2 fa5cf58 fee2e0a fa5cf58 fee2e0a 034b2f2 2f5c91e 034b2f2 c11c555 2f5c91e c11c555 034b2f2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
# app.py
import gradio as gr
from transformers import BlipProcessor, BlipForConditionalGeneration
from gtts import gTTS
import io
from PIL import Image
# -------------------------------
# Load BLIP-base model (lighter version)
# -------------------------------
processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
# -------------------------------
# Generate caption function
# -------------------------------
# def generate_caption_tts(image):
# caption = generate_caption(model, processor, image)
# audio_file = text_to_audio_file(caption)
# return caption, audio_file # return file path, not BytesIO
# -------------------------------
# Convert text to speech using gTTS
# -------------------------------
import tempfile
import pyttsx3
def text_to_audio_file(text):
# Create a temporary file
tmp_file = tempfile.NamedTemporaryFile(suffix=".mp3", delete=False)
tmp_path = tmp_file.name
tmp_file.close()
engine = pyttsx3.init()
engine.save_to_file(text, tmp_path)
engine.runAndWait()
return tmp_path
# -------------------------------
# Gradio interface: Caption + Audio
# -------------------------------
def generate_caption_tts(image):
caption = generate_caption_from_image(model, processor, image) # uses global model/processor
audio_file = text_to_audio_file(caption)
return caption, audio_file
interface = gr.Interface(
fn=generate_caption_tts,
inputs=gr.Image(type="numpy"),
outputs=[gr.Textbox(label="Generated Caption"), gr.Audio(type="filepath", label="TTS Audio")],
title="Image Captioning for Visually Impaired",
description="Upload an image, get a caption and audio description."
)
interface.launch()
|