Spaces:
Runtime error
Runtime error
File size: 1,730 Bytes
a761387 eaa3654 cf2cafe a761387 cf2cafe 6919033 eaa3654 6919033 6b4a9a6 829b25c eaa3654 829b25c 6d97bc1 829b25c 5fe2fff 829b25c 6919033 cf2cafe 6919033 829b25c 8f2fa48 a761387 8f2fa48 |
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 |
# app.py
from flask import Flask, render_template, request, jsonify
from transformers import pipeline, BlipForConditionalGeneration, BlipProcessor
import torchaudio
from torchaudio.transforms import Resample
import torch
import gradio as gr
app = Flask(__name__)
# Initialize TTS model from Hugging Face
tts_model_name = "Kamonwan/blip-image-captioning-new"
tts = pipeline(task="text-to-speech", model=tts_model_name)
# Initialize Blip model for image captioning
model_id = "Kamonwan/blip-image-captioning-new"
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('/generate_caption', methods=['POST'])
def generate_caption_api():
image = request.files['image'].read()
generated_caption, audio_path = generate_caption(image)
return jsonify({'generated_caption': generated_caption, 'audio_path': audio_path})
@app.route('/')
def index():
return render_template("index.html")
if __name__ == '__main__':
app.run(host='0.0.0.0', port=5000)
|