app.py

import pathlib
import uuid
import os
import gradio as gr
from tqdm import tqdm
import requests
import urllib.request
import json
import time
output_mp3="output.mp3"

def upload_image(img: str, d_id_key: str):
    url = "https://api.d-id.com/images"

    files = {"image": ("hero.jpg", open(img, "rb"), "image/jpg")}
    headers = {
        "accept": "application/json",
        "authorization": "Basic "+d_id_key
    }

    response = requests.post(url, files=files, headers=headers)

    response_dict = response.json()
    img_url = response_dict["url"]
    # return audio_url
    print(img_url)
    return img_url

def upload_audio(audio: str, d_id_key: str):
    url = "https://api.d-id.com/audios"

    files = {"audio": (audio, open(audio, "rb"), "audio/mpeg")}
    headers = {
        "accept": "application/json",
        "authorization": "Basic "+d_id_key
    }
    response = requests.post(url, files=files, headers=headers)
    response_dict = response.json()
    audio_url = response_dict["url"]
    # return audio_url
    print(audio_url)
    return audio_url

def get_did_video(process_video_url,d_id_key):
    url = "https://api.d-id.com/talks/"+process_video_url

    headers = {
        "accept": "application/json",
        "authorization": "Basic "+d_id_key
    }
    response_dict = {}

    while "result_url" not in response_dict:
      # make API call and get response dictionary
      response = requests.get(url, headers=headers)
      response_dict = response.json()
      
      print(response.text)
      
      # wait for 1 second before checking again
      time.sleep(1)
      
    # "result_url" key is now present in the dictionary
    result_url = response_dict["result_url"]

    print("From did_video \n\n\n")
    print("/n/n/n")

    # response_dict = response.json()
    result_url = response_dict["result_url"]

    print(result_url)
    return result_url

def text_to_speach_api(text: str, elv_key,voice_id: str):
    url = "https://api.elevenlabs.io/v1/text-to-speech/"+voice_id+"/stream"
    headers = {
        "accept": "*/*",
        "xi-api-key": elv_key,
        "Content-Type": "application/json",
    }
    data = {
        "text": text,
        "voice_settings": {
            "stability": 0,
            "similarity_boost": 0
        }
    }

    response = requests.post(url, headers=headers, json=data)
    # print(response.text)

    if response.ok:
        with open("output.mp3", "wb") as f:
            f.write(response.content)
    else:
        print("Error: ", response.text)

def get_voice_names():
    with open("data.json") as f:
        data = json.load(f)
    return [voice["name"] for voice in data["voices"]]



# define a function to get voice id by name
def get_voice_id(name):
    # load the JSON data
    with open("data.json") as f:
        data = json.load(f)
    for voice in data['voices']:
        if voice['name'] == name:
            return voice['voice_id']
    return None

#D-id API
def d_id_api(image_url, d_id_key,audio_url):
    print("D-id API")
    url = "https://api.d-id.com/talks"
    payload = {
        
    "source_url": image_url,
    "script": {
        "type": "audio",
        "audio_url": audio_url,
    }

    }
    headers = {
      "accept": "application/json",
      "content-type": "application/json",
    "authorization": "Basic "+d_id_key
  }

    response = requests.post(url, json=payload, headers=headers)
    print("From D-id API \n\n\n")
    print(response.text)
    response_dict = response.json()
    process_video = response_dict["id"]
    # return audio_url
    print(process_video)
    return process_video
    


def transcribe_video(d_id_key: str, elv_key: str, full_text: str,voice_name: str,img):
    print(voice_name)
    voice_id=get_voice_id(voice_name)
    text_to_speach_api(full_text, elv_key,voice_id)
    audio_url=upload_audio(output_mp3,d_id_key)
    image_url=upload_image(img,d_id_key)
    process_video_url=d_id_api(image_url, d_id_key,audio_url)
    video_url=get_did_video(process_video_url,d_id_key)
    file_name = 'hero.mp4'
    urllib.request.urlretrieve(video_url, file_name)
    return file_name
    

examples = [["", "","Good morning, it's great to see you! I hope you're having a wonderful day. I just wanted to say thank you for taking the time to speak with me. Is there anything new or exciting happening in your life? I'd love to hear about it. Let's catch up soon!",
             "Arnold","./images/hero.jpg"],["","","Hello there, I'm a talking photo! I can speak any text you type here. Try it out!", "Domi","./images/3.jpg"],["","","Hello there, I'm a talking photo! I can speak any text you type here. Try it out!", "Domi","./images/2.jpg"]]

demo = gr.Interface(fn=transcribe_video, inputs=[
    gr.Textbox(label="D-Id API Key",placeholder="Paste your D-Id",type='password'),
    gr.Textbox(label="Elevenlabs API Keys",placeholder="Paste Elevenlabs",type='password'),
    gr.Textbox(lines=4, label=" Please input the text you wish to generate in order to make the photo speak.", placeholder="English Text here"),
    gr.Dropdown(choices=get_voice_names(), label="Select a voice"),
    gr.Image(label="photo of a Person", type="filepath")
  ], outputs="video",title="Bring your images to life with the talking animation feature now!",examples=examples,cache_examples=False)

demo.launch()