app.py

import gradio as gr
from PIL import Image  
import re
import os
import speech_recognition as sr


stable_diffusion = gr.Blocks.load(name="spaces/stabilityai/stable-diffusion")
r = sr.Recognizer()

def transcribe(audio):
    with sr.AudioFile(audio) as source:
        audio_ = r.listen(source) 
        text = r.recognize_google(audio_)#, language = 'en-IN')# , show_all=True)
    return text



def get_images(prompt):
    gallery_dir = stable_diffusion(prompt, fn_index=2)
    return [os.path.join(gallery_dir, img) for img in os.listdir(gallery_dir)]

with gr.Blocks() as demo:

    gr.Markdown("Stable diffusion magic -> Get the photo from whatever you can think of!")
    with gr.Tab("Audio Input"):
        audio_input = gr.Audio(source="microphone", type="filepath")
        submit_audio_button = gr.Button("Convert to Image")
        text_output = gr.Textbox(label="Recorded text")
        
    with gr.Tab("Text Input"):
        text_input = gr.Textbox(label="Enter text")
        submit_button_text = gr.Button("Convert to Image")


    
    # output = gr.Textbox(label="Output Box")
    sd_output = gr.Gallery().style(grid=2, height="auto")

    submit_audio_button.click(fn=transcribe, inputs=audio_input, outputs=text_output)
    text_output.change(fn=get_images, inputs=text_output, outputs=sd_output)
    submit_button_text.click(fn=get_images, inputs=text_input, outputs=sd_output)


demo.launch()