app.py

from flask import Flask, request, jsonify
from transformers import pipeline, BlipForConditionalGeneration, BlipProcessor
import torchaudio
from torchaudio.transforms import Resample
import torch
from io import BytesIO

app = Flask(__name__)

# Initialize TTS model from Hugging Face
tts_model_name = "suno/bark"
tts = pipeline(task="text-to-speech", model=tts_model_name)

# Initialize Blip model for image captioning
model_id = "dblasko/blip-dalle3-img2prompt"
blip_model = BlipForConditionalGeneration.from_pretrained(model_id)
blip_processor = BlipProcessor.from_pretrained(model_id)

def generate_caption(image):
    # Generate caption from image using Blip model
    inputs = blip_processor(images=image, return_tensors="pt")
    pixel_values = inputs.pixel_values
    generated_ids = blip_model.generate(pixel_values=pixel_values, max_length=50)
    generated_caption = blip_processor.batch_decode(generated_ids, skip_special_tokens=True, temperature=0.8, top_k=40, top_p=0.9)[0]

    # Use TTS model to convert generated caption to audio
    audio_output = tts(generated_caption)
    audio_path = "generated_audio_resampled.wav"
    torchaudio.save(audio_path, torch.tensor(audio_output[0]), audio_output["sampling_rate"])

    return generated_caption, audio_path

@app.route('/upload', methods=['POST'])
def upload_image():
    if 'image' not in request.files:
        return jsonify({'error': 'No image provided'}), 400

    image_file = request.files['image'].read()
    generated_caption, audio_path = generate_caption(image_file)

    return jsonify({'generated_caption': generated_caption, 'audio_url': audio_path}), 200

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000, debug=True)